Eight Lessons Learned In Edinburgh
Before implementing Logan in Glasgow it is worth internalising some of the hard-won lessons and insights from the slow, painful Edinburgh experience.
The Edinburgh institutions listed by Logan did not emerge like Pallas Athena, fully armoured from the forehead of Zeus, but grew, part organically, part experimentally and part consciously.
And critically they grew out of each other.
Informatic Ventures first begat the OG social forum, the StartUpCafe website and then begat TechMeetUp — the first and longest lasting open-tech and startup agora in Scotland. Without something as simple as an agora, an open forum, a meeting place were different techies and entrepreneurs met, chatted and connected, then there would have been nothing.
The offspring continued. TechMeetUp sprogged in turn the Turing Festival (and the Turing Fringe), and their children include CodeBase (via TechCube) and the modern infrastructure.
TechMeetUp also metastasised. The Scottish Lean Circle came out as a parallel meetup with a business focus, and was superseded and replaced by Product Tank. Turing Festival, which was first at tech-aligned, migrated to the full company building position it now occupies.
Informatics Ventures also launched and curated the EIE investment event.
Modern Edinburgh tech is the great grandchild, the product of evolution, of selection, of success promoted and failure analysed.
This matters is because TechMeetUp is not, and never was, an Edinburgh thing. Bizarrely TechMeetUp started at a Y Combinator meetup in London and a trade show in the Excel conference centre.
Within 2 or 3 months Edinburgh TechMeetUp was joined by the Glasgow chapter and later by Aberdeen, Dundee, Stirling and Inverness.
Edinburgh flourished and Glasgow didn’t. The purpose of this report is try and explain why, to identify the missing nutrients and irrigation that the emerging Edinburgh scene got and Glasgow did.
This failure was not because the Glasgow scene was less ambitious, talented or hard working. Let me single out Rookie Oven, CodeYourFuture and TechMeetUp Glasgow here (I am sure there are others I haven’t met).
Informatics Ventures were the Logan Report *avant la lettre*. In many ways this entire document is a detailed history of years of patient work — and due respect to Andrew Mitchell and Danny Helson and the rest of the team there.
Lets look in detail at how the Edinburgh scene was built, the thought processes, the experiments, what worked and what didn’t — from the inside. I had a couple of startups in 2005 and 2007, spoke at the first TechMeetUp, started the Scottish Lean Circle and co-founded the Turing Festival. My cofounder went on to start CodeBase.
We begin with lessons from Edinburgh.
Lesson 1 — The Harlem Globetrotters Problem
Edinburgh had to address the Harlem Globe Trotters question. Everyone wants to be Silicon Valley, everyone is studying their playbook. But like the Harlem Globetrotters, if you go head to head with Sand Hill Road you will lose. They are better at being themselves than you can ever be. Copying San Francisco directly is crazy.
In Edinburgh we consciously chose another way. The only way to beat the Harlem Globetrotters is: don’t play basketball.
Edinburgh Tech is not organised like Silicon Valley Tech, inspired by, yes and aspiring to, yes, but organised like? No. Our institutions, companies and their relationships are indigenous, organic and quite, quite different.
Startup founders are trained to think of network effects, of moats and protected positions, of Metcalfe points. Edinburgh applied company-building to ecosystem building, systematically. Edinburgh would out-compete London on the quality and density of its networks and the integration of its ecosystem.
Early on — when there were only about 10 startups, Scottish Enterprise took a group of founders out to Silicon Valley. (It remains the single most transformative thing you can do.) During the trip Google poached one of the founders with a good salary and place on their Product Manager’s training programme. Jennie Lees took it, understandably. And then we were 9.
This seemed an insurmountable problem — the tech scene should simply evaporate. You could double your money in London, treble it in Cali, why wouldn’t you? With that came two revelations:
- big Silicon Valley firms saw strong academic computer science Universities as recruitment sources — we should focus on our success not theirs.
- people leave for money and opportunity, its simple. Our focus should be on the complex and nuanced issue of retention.
Asking people around the scene why they didn’t leave? the Meadows, walking to work, love, Auld Reekie herself, friends, comfort, loyalty to their startup and co-workers. Human factors all, and these cheerfully co-existed with global scale ambition, but also with doubt that it was possible to achieve that.
We should focus on why and how companies were successful in California and create distinctive organisational solutions in Scotland that would create the same outcomes by different means.
Silicon Valley pumps out stuff about young founders and college-age whizz kids, yet experienced founders in their 30s and 40s make more exits. They see company creation as frog-spawn. Create as many as possible, don’t worry too much about each individual one, but back every froglet that makes it out of the pond. You plant them in the rich compost heap of failed startups.
If you hire 5 new-out-of-college-kids in Paolo Alto, he was CEO at his undergraduate startup, she did Product/CTO, they had a lecture course from Mark Zuckerberg in their 2nd year. Behind them in serried ranks are ex-Google, ex-Facebook, ex-Amazon people in their 20s, in their 30s. Recruitment, staff development, enskilling, no problem, hire and learn on the job.
By contrast our start up scene was intrinsically mammalian. We had gey few babies and when the Google eagle swooped off with one we all grieved: each child, each employee is precious. Running our companies like SF startups was a recipe for disaster.
In a skills-poor environment training is critical — and the core mantra of in-company training in Edinburgh is don’t do it. The model of networking, of meetups, of product tank, or the Turing Festival is extra-company training. Don’t keep training and learning to yourself, organise a meetup to do it on a community wide basis.
The old mantras of the Scottish Lean Circle were no competitors on this side of the Atlantic and when somebody is presenting, everybody in the room works for their company. Scotland cannot afford to wait for your company either to go bankrupt or be successful enough to slough off senior staff to learn from your successes and mistakes. Nor can your company afford to hire in the skills and learning you need, and if they could you could not find them. Everybody needs to learn from everybody else’s experience at the same time.
One of the core underpinnings of this philosophy is the aggressive externalising of training — the sharing of experience. In Silicon Valley it is trivial to pick up a team of experienced individuals and in Scotland it is not. American firms typically focus on internal training.
In a country without the scale this is a bad idea. Much better to externalise all growth in capability by starting meetups, and then externalising training to community-based conference.
The Turing Festival emerged from a mass of lower-level community wide trainings, meetups and happenings:
Critically the speakers and attendees and organisers were all wearing many hats: as employees in their firms, as organisers of meetups, as chairs, as speakers.
As the scene grew Turing grew with the big companies and tracked their needs and radically changed itself.
Today Turing remains unique amongst major tech events in its close organic relations with its local tech scene all the way up the funnel from pre-company formation to unicorn.
As Logan identifies the market position placed pressure on Turing to charge commercial rates. But the community roots of Turing have been maintained. The event is staffed from the startup community, students and employees of early stage companies.
Staffing is on a 1-day work/1-day attend basis with full access to the conference social events, fringe meetups and parties and networking opportunities.
Lots of the stewards return year on year. Working-at-Turing is designed as the Tinder-for-new-founders — an important community resource in its own right.
Turing also has excellent outreach facilities including pre-emptive training for would-be-volunteers on how to maximise the networking and communication opportunities of volunteering at Turing.
Lesson 2 — Externalised Excellence
The Harlem Globetrotters lose on our pitch. It is not an accident that the Turing Festival is Festival and not a conference. It is not a coincidence that it takes places immediately before the Edinburgh Festival (in the beginning it was during the festival).
It is not a mistake that the first class tech speaker, Steve Wozniak — the co-founder of Apple with Steve Jobs — spoke at the distinctly 5th rate Turing Festival. He came for the Edinburgh Festival.
Any proposal that involves international travel, financiers, yada-yada, that doesn’t also involve golf, whisky, sailing and the shortbread tins of the Highlands and Islands is missing a trick.
Successful wooing of global companies, or external investors, must build on Scotland-the-myth.
Lesson 3 — Heartbeat Meetups
The density depth and interconnectedness of the Edinburgh tech community is driven by the structure of its meetups.
At the core is a heartbeat meetup with a core mailing list, the sun, the source of light and life, and orbiting around it are transactional meetups.
The organisers, key members and local speakers at the moons are overwhelmingly members of the core heartbeat community itself.
TechMeetup has an eclectic record of talks ranging from home made robots, deep tech, the politics of tech, toy products, new programming languages, lost paradigms and more. It is a showcase for new speakers, new ideas and new communities. It has the “intermittent reward” that makes Twitter so addictive. Sometimes its absolutely amazing but there is no way of knowing when.
Edinburgh TechMeetup had several major advantages over Glasgow:
- it was located in the incubator floor at Appleton Tower, alongside the startups including what become FanDuel
- it had an institutional sponsor (Informatic Ventures) which had a strategic interest in building the community long-term
- a regular sponsor meant beer and pizza, every time
The importance of regular supplies of free food and drink cannot be over-emphasised.
Early stage startups are underpaid, anti-social and have gruelling working hours. In Edinburgh you could (and still can) get food, drink and some social life 4 or 5 nights, free.
As it costs about £5K to hire an engineer, sponsorship of meetups is a cheap and cheerful KPI for community health. Tracking what companies fund what communities and meetups with what amounts yields a surprising amount of insight.
A strong heartbeat organisation is a key component of anti-fragility. Meetups are volunteer-led and are subject to the vicissitudes of the volunteer life: sickness, work, burn-out and love. Having a well-funded and resilient heartbeat means that the the side meetups can come and go, fall into desuetude, hibernate and then wake up.
The Erlang/BEAM community is a good example with several bursts of activity, a important conference at Turing and long gaps before being revived by new generation of Elixirists independently.
Without a heartbeat meetup to use to relaunch, each iteration would have been incredibly painful.
Lesson 4 — The Scotch Pivot
At the 2nd Turing Festival we had a long battle over whether there should be a session on “billion dollar babies” — the roadmap to Scottish Unicorns. Simply voicing that ambition was an achievement in itself. Nigel Eccles from FanDuel and Gareth Williams from Skyscanner spoke — both on the road to, but far off, unicorn status then.
Nigel talked about the extra-ordinary step he had just taken. His team had raised a huge sum, half a million, for HubDub, a news aggregator, launched a product to rave reviews with strong user interest and then closed that down and pivoted to an entirely new sector, and totally different product — FanDuel. He explained his reasoning with a slide like this — the famous “Scotch Pivot”.
A business model which is fundable in the US is not fundable in Scotland, and won’t be for the foreseeable future. The investment cashflow is not strong and steady enough to support high-capital, slow-burn companies.
Hence the proverb American CEOs raise money from VCs, Scottish Product Managers raise money from customers.
It profoundly altered the structure of the Edinburgh Tech Community.
Lesson 5 — The Double Heartbeat
As a consequence of the Scotch Pivot talk the Edinburgh scene underwent a major transformation to a double-heartbeat model. A new heartbeat, business focussed event — first the Scottish Lean Circle and then Product Tank — with its own associated communities and networks was developed.
Lesson 6 — Sequencing And Repetition
A complex event calender needs to be managed.
An infrastructure was built to handle the sequencing and management of multiple events. FanDuel sponsored the development of the Open Tech Calendar, considerable thought was given to sequencing and integration, the programme of heart beat events across Scotland was managed, and critically Turing emerged as part of this infrastructure. Shout out to James Baster, a community stallwart for his work on it.
Repetition in community building has considerable power: Constant reiteration of, and grooming of networks helps prevent them getting link rot as people leave for love, work or simply a different life.
Lesson 7 — Roads To Global
American firms have substantial advantages in the road to global presence. They have a continental scale domestic market and dominating that domestic market is the first priority. It provides the financial and funding firepower to curb stomp or purchase competitors in other countries.
There are successful EU business models where companies are built-to-be-purchased by the US market leader in a sector — a clone-and-flip model. (There is also the successful Israeli model of building companies slightly ahead of the market in niche infrastructural technology to sell them off to US would-be-hegemons — build-to-sell).
Wannabe global firms in Europe have to chose their own roadmap to the world. As FanDuel established itself as a competitor to Yahoo sports it chose one way. The HQ/business side of the team moved to New York taking a fancy office directly across from Tammany Hall. The engineering side stayed in Edinburgh.
By contrast both Skyscanner and bet365 chose a different route:
bet365 were constrained because American gambling is mobbed up. The Mafia take a dim view of competition-through-innovation and simply bought political influence to make online gambling illegal.
A US resident using a credit card to bet online has led to British betting executives being taken off diverted planes and jailed.
The bet365 approach is simply to ban Americans, or people who might possibly be Americans from betting.
bet365 is a veritable United Nations. When I left my flat in the Etruria estate (where bet365 owned half the houses) and walked to work along the canal I was swept along in a river of foreign tongues: greek, bulgarian, spanish, italian, portuguese, chinese, korean as the employees necessary to support a business in 80 languages and as many legal jurisdictions.
It is instructive to visit the Skyscanner office up at Simpsons on the Meadows. There is a long mural/chart on the ground floor showing the company’s growth. When Skyscanner had 5 employees it was available in 6 languages. The road to making Europe a continental scale market goes through become multi-lingual very, very quickly.
An ecosystem that contains multitudes, as Edinburgh currently does, is a dramatic help.
There is a 3rd road to global which no Scottish company has followed — the Commercial Open Source Software (COSS) model. This model sells support to other tech businesses. The internal language of tech is Standard International American Business English.
The COSS path has its own phasing model with essentially building out support first in one timezone (EU or US) and then adding another 2 for follow-the-sun support. Typically the workforce is fully remote.
Lesson 8 — The Success And Failure Of The Turing Fringe
The modern Turing Festival is different from its initial design which had two parts like the Edinburgh Festivals:
It was designed with a curated, high end Festival and a give-it-a-go Fringe. Turing as we know it evolved out of the curated end and the Festival Fringe petered out.
The failure of the Fringe was an organisational and not a financial matter — the year the main Festival was at most risk, when it lost a lot, the Fringe made a healthy profit and cross-subsidised it.
The design of the Fringe came from applying the theory and practice of network effects to everything, not just product marketing. The Edinburgh Festival Fringe holds its position against a million wannabes because it is a classic two-sided market: punters and shows. This cross-network protection resembles the dominant two-sided positions that LinkedIn and eBay still maintain.
Turing was explicitly named and aimed at becoming the European version of South-By-South-West, explicitly programmed like a musical festival, with multiple stages and performers ranging from international superstars to, basically, buskers.
Failure of the Fringe was its inability to become self-programming. At its best it had 5 or 6 different events and it never managed to become an exploding mass of conferences. (During the Festival Edinburgh has a grave shortage of venues suitable for art, performance or comedy, but maintains a surplus of venues suitable for tech events ie lecture rooms with desks).
One of the major financial underpinnings of the modern commercial fringe is licensed premises but the major comedy players have exclusive rights from Edinburgh University. So Turing was locked out of this lucrative market (not for want of investigating).
The primary surplus at the Fringe comes from recruitment costs — companies are happy to pay for access to talent. In this context acquiring sponsorship money is comparatively easy.
The Turing Fringe resembled Andrew Neil’s Sunday Times in its pomp — a food section, a travel section, a personal finance section, a motoring section — each aimed at, and tailored for a different advertising market.
The Fringe model has at its heart an anti-fragile mechanism. In the curated model, financial risk is with the festival, it flies in the speakers, it rents the venue and if sales and sponsorship fail then goes bust.
With a Fringe model the financial risk is put on the external parties that organise events under its umbrella. They pay their speakers (or not), spend on travel (or not) and pay subs to the Fringe Org for marketing/venues/central services.
The primary reason for the failure was organisational — having a Fringe organiser who had a full time job to recapitalise his failing startup (so a full time side job as well) and was mostly living in Stoke-On-Trent was not sustainable.
