Speculative thoughts on a 4th Generation Accelerator

[drllau]

The Australian govt has just released a report on their startup/accelerator scene. I spent a decade in academia in Brisbane and still many fond memories and keep a hand in their Silicon Beach forum. As result ended up in an email discussion with Martin (author of report) a while back, the gist of which was

I've had discussions with visiting VCs and one recent trend is to add value by acquiring persistent resources (like HR) which their portfolio companies don't waste) ... there are structural issues mainly on the financing side and with the various angel platforms out there, there is simply no proprietary deal-flow any more. So instead you end up with a beauty contest model (or to be charitable think of Aussie Rules competition from varsity to pro) where the accelerators pick and groom the best prospects for investors to look over. So they are paying up to 50% (rest goes to the startup) for the management and their mentors. See https://groups.google.com/forum/#!topic/openfrog/DE9NZ7Jh-T0 on verticals and my observations on specialisation (I look at innovation systems as academic curiosity to my day job as IP broker). 

So whilst I may be somewhat cynical using phrase "beauty contest", accelerators play a role in the capitalist scene of allocating resources, in this case the most important one is the time/attention of serious investors, whether angel or corporate. You can read the report for the fine details (bearing in mind Australia is closer to a blended family of squabbling siblings compared with the little treasure of Singapore) I'd like to toss out for discussion what would be the next generation of accelerators (independent of investment stage).

Accelerators can be seen to morph through 3 generations or economic models

- cost reduction - typically landlord/tenant relationship where co-sharing or clustering allows for reduced costs like rental

- convenient - addition of value-added services like accountants or mentors (beyond the mates of owner)

- connective - referral to next stage or helping establish networks, either suppliers, potential M&A (ie accuhire) or business partners

So what would be the next evolution? As accelerators gain staying power (I'd guess minimum decade) they'd accumulate resources and effectively become a nexus but exactly what are the success factors? Technology changes rapidly so the supporting infrastructure also has to adapt (Darwinian pressure). So what would a 4th gen accelerator look like in the next decade? Some speculative thoughts organised in decreasing likelihood

- become mainstream business development process ... the sheer pace of tech means that the innovative absorptive capacity of individual firms or even MNCs won't be sufficient ... one direction is sectorial accelerators like auto industry for autonomous vehicles or reversion to state control (cf Russian grooming model for Olympic)- control of specialist resources ... not just office space like JTC in Block 71 but unique ... I'm contemplating the ShanZia flex menufacturing in Shenzhen or perhaps rapid deployment space launch capabilities

- unique culture ... when looking at Millenials, their concept of career would be different to the tradition work ladder or corporate hierarchy. Simply the misfits (cf skunkwork) cannot survive in a corporate profit centre (unless protected by sponsor) so there probably needs to be a mechanism to spin in/out these new ideas but the eco-system needs both sustaining talent and a meme. A good analogy is the Hollywood actress/waiter Cinderella stories ... basically a mental meme to encourage a group of fanatics to go through lean times for the sake of success (however defined including non-profit).

I've some more tentative ideas but I'd leave space here for people to toss in how they see accelerators morphing

Lawrence

http://www.linkedin.com/in/drllau

[Wong Meng Weng]

On Thursday, June 9, 2016 at 8:26:11 PM UTC+1, drllau wrote:

So what would be the next evolution? As accelerators gain staying power (I'd guess minimum decade) they'd accumulate resources and effectively become a nexus but exactly what are the success factors? Technology changes rapidly so the supporting infrastructure also has to adapt (Darwinian pressure). So what would a 4th gen accelerator look like in the next decade? Some speculative thoughts organised in decreasing likelihood

Some accelerators notably YC and TechStars have managed to stay independent and profitable on exits. However, as the volume and value of exits in SEA do not compare to the Valley, accelerators in SG are unable to reach breakeven on exits alone.

This tends to lead to a pivot toward corporate acceleration, in which incumbents get involved with startup innovation through a sponsorship or client/vendor relationship with an accelerator.

http://www.afr.com/technology/enterprise-it/pollenizer-swaps-startups-for-big-names-20140922-jftxn

As with well-established institutions like Harvard, Stanford, Penn, etc, the network effects of any given accelerator's alumni community grow with size. In Singapore, the proliferation of accelerators has limited the value of any one alumni network; YC by contrast has over 1000 alumni in its network, creating "winner takes all" network effects. It would be an interesting experiment to contrast another ecosystem which decided to put more wood in fewer quivers.

[drllau]

Y-combinator tries to optimise home-runs (unicorns) rather than hit-rates (#survivors) so it may not be typical. As the report noted, there is a philosophical difference between buffer accelerators (typically govt funded like Startup Chile) v bridging accelerators with a strong focus on skills/bizDev on the former and networks in the latter. There is also a more recent observation of mentor fatigue. Just because you have a few well-known stars, how much practical interaction time is there will the mentors? 

>This tends to lead to a pivot toward corporate acceleration, in which incumbents get involved with startup innovation through a sponsorship or client/vendor relationship with an accelerator.

http://www.afr.com/technology/enterprise-it/pollenizer-swaps-startups-for-big-names-20140922-jftxn

This is due to my observation#3 in that the younger crowd are just not as attracted to traditional corporate hierarchies and those incumbants are facing existential threats without adjusting their innovation policies ... Detroit v Telsa, Circuit City v eBay, etc and now potentially FinTech. As such corporate accelerators are just one mechanism to try and adopt the DNA of startups with varying degrees of success (see P&G with their Innocentive hub+spoke bounty system or SiliconValley outposts). Remember that if the definition of accelerator is retained knowledge, this tacit knowledge tends to crystalise in different ways, eg specific training regimes or improved selection criteria. Depending on industry, some of it can cross-fertilise.

However, I try to look beyond the existing status quo and conjecture what is missing. In the general scheme of cost-saving -> convenience/concentration - connectivity/community my hunch is that there is missing catalyst/coordination role. The reason is two-fold. Most accelerators only accept early stage firms with typically <5 founders. As such their selection criteria usually prefers established working relations,or existing teams. However, there's an observation that on NY theatre scene the most successful Broadway shows have a 2:1 ratio of existing troup + newblood. This is missing with accelerators in that there is no deliberate mixing of teams within the cohort (unlike Startup Weekend) so idealation/creativity is suboptimal IMHO. Also with new tech, it takes time to figure out the new operating modalities ... for example the early web-sites were essentially translation of mail catalogs until eBay reframed the eCommerce space. There's this lack of critical thinking about how new tech (like blockchains) can create new business models so the strenght of mentor networks (with prevalent logic) may actually be a hinderance.

> In Singapore, the proliferation of accelerators has limited the value of any one alumni network;

This is govt policy to allow for competitive ecosystem ... but there are mechanisms to overcome this like cross-cutting DemoDays or exchange programs or regional feeder networks.  But in terms of next generation of accelerators, what would surpass the existing system?

[Hugh Mason]

On Thursday, 9 June 2016 21:26:11 UTC+2, drllau wrote:

I've had discussions with visiting VCs and one recent trend is to add value by acquiring persistent resources (like HR) which their portfolio companies don't waste)

Meng contributed greatly to my Masters Thesis which I wrote at a time when it was a lot less clear what an 'accelerator' actually is and how it differs from an 'incubator' - it turns out that there's quite a lot of relevant literature which I reviewed here: http://www.academia.edu/11662129/Guilds_for_Geeks_-_Describing_Business_Acceleration_using_Grounded_Theory_Methods

Since I completed that thesis Susan Cohen has written far better than I ever could on accelerators: http://www.mitpressjournals.org/doi/pdf/10.1162/INOV_a_00184 and I recommend googling for her growing library of papers on the subject

 

... there are structural issues mainly on the financing side and with the various angel platforms out there, there is simply no proprietary deal-flow any more. So instead you end up with a beauty contest model (or to be charitable think of Aussie Rules competition from varsity to pro) where the accelerators pick and groom the best prospects for investors to look over. So they are paying up to 50% (rest goes to the startup) for the management and their mentors. See https://groups.google.com/forum/#!topic/openfrog/DE9NZ7Jh-T0 on verticals and my observations on specialisation (I look at innovation systems as academic curiosity to my day job as IP broker). 

One of the things I came to realize through writing my thesis is that talking about accelerators is often rather like the proverbial elephant being described by five blind men. So someone looking from a macroeconomic perspective might see it as described in your paragraph above but another stakeholder could see it very differently. The practical experience Meng and I have had running the first accelerator in the region is that it's best understood as a social construct, or in economic terms as a platform business where many of the stakeholders (eg mentors) have a non-financial transaction with the accelerator.

The folk who try to simplify an accelerator down to a process - a kind of carwash for entrepreneurs - miss all that and this is why the mentor fatigue etc comes in because they don't build the structures to create the non financial return that mentors are looking for. It's easy to go round the world and look for best practice and to try to plant it in your own garden but whether or not it takes root is going to be very hit-and-miss. So the risk in an environment like Singapore, where so many transplants have taken place without much reflection or discussion of the fundamental mechanisms, is that we are making the classic startup mistake of premature scaling.

 

So whilst I may be somewhat cynical using phrase "beauty contest", accelerators play a role in the capitalist scene of allocating resources, in this case the most important one is the time/attention of serious investors, whether angel or corporate. You can read the report for the fine details (bearing in mind Australia is closer to a blended family of squabbling siblings compared with the little treasure of Singapore) I'd like to toss out for discussion what would be the next generation of accelerators (independent of investment stage).

The key question here is - to serve what purpose? Investors, startups and mentors don't think top-down macroeconomics. And the view of many investors that accelerators are there to 'feed' them is only one perspective. Certainly entrepreneurs don't see their role as being to serve investors - rather the other way around. Hence the very first thing that Brad Feld and David Cohen stressed when they spoke to Meng and I back in 2010 - the entrepreneurs are the only people that really create wealth in this picture. Everyone else is a servant or an enabler/block.

So my observation is that the most successful folk in the accelerator business take a community perspective, which is why I ended up using one of Meng's phrases at the title for my thesis - accelerators are 'guilds for geeks'. They need to be built bottom-up not top-down to suit the context in which they are intended to flourish and that means their form will be contingent on the context, much as flowers and plants look very different in different contexts. Taking that analogy a bit further - plants do not evolve to feed the herbivores that eat them and I personally don't believe there is a role for some great ecosystem engineer in the sky interfering at every level from the Krebs cycle upwards.

Accelerators can be seen to morph through 3 generations or economic models

- cost reduction - typically landlord/tenant relationship where co-sharing or clustering allows for reduced costs like rental

- convenient - addition of value-added services like accountants or mentors (beyond the mates of owner)

- connective - referral to next stage or helping establish networks, either suppliers, potential M&A (ie accuhire) or business partners

You will find this model of incubator development referenced in my thesis. 

Thinking of what the next generation of accelerators might be, it could be helpful to be very clear about the differences between incubators and accelerator which, again, Susan Cohen (with Yael Hochberg) describe very well (and which is reported here in HBR) : https://hbr.org/2016/03/what-startup-accelerators-really-do

[drllau]

On Friday, 10 June 2016 13:12:24 UTC+8, Hugh Mason wrote:

> The practical experience Meng and I have had running the first accelerator in the region is that it's best understood as a social construct, or in economic terms as a platform business where many of the stakeholders (eg mentors) have a non-financial transaction with the accelerator. 

>So my observation is that the most successful folk in the accelerator business take a community perspective, which is why I ended up using one of Meng's phrases at the title for my thesis - accelerators are 'guilds for geeks'.

This fits into my #3 which I lump as "culture". In this regards Singapore has outdone its neighbours inculturing a more risk taking attitude in current generation compared with the traditional govt officework or MNC salaryman. Because there are too many spill-over effects, the govt will still need to provide an strong input unlike the more Laissez-Faire SiliconValley.  Hence Singapore is probably more sensitive to policies like migration and tax-risk tradeoffs. On the otherhand, I suspect the time is ripe to do more serious social engineering. SG has gone from the 80s model of operating uni-level singles club and I'm aware some accelerates have specific community managers to foster the "right" culture. Unfortunately unlike mobileApps, it's hard to do A/B testing but some serious thought in social engagement and what works might be one of the hallmarks of 4th gen accelerators (cf old Sandhill charity events). 

 

>

One of the things I came to realize through writing my thesis is that talking about accelerators is often rather like the proverbial elephant being described by five blind men. So someone looking from a macroeconomic perspective might see it as described in your paragraph above but another stakeholder could see it very differently. 

So the risk in an environment like Singapore, where so many transplants have taken place without much reflection or discussion of the fundamental mechanisms, is that we are making the classic startup mistake of premature scaling.

The key question here is - to serve what purpose? Investors, startups and mentors don't think top-down macroeconomics. And the view of many investors that accelerators are there to 'feed' them is only one perspective.

Perhaps a simple analogy is to use a fire = fuel (financial capital) + oxygen (ideas) + spark (entrepreneurs). You need the right combination and depending on mix, can either be slow burn or massive explosion. What has happened since 90s is that the fuel source has gone from lumps of coal (concentrated in places like SV or NY/London) as financial capital is now internationally mobile. What people are realising is that the oxygen level has also gone up as the social capital in the form of free/open software, PaaS, toolchain platforms is allowing startups to be created anywhere. The problem is how to encourage sparks (the entrepreneurs no matter their age) to ignite, to arise from the ashes of failure to recall the flames of passion. Here I think Singapore probably overemphasise youth as the most inspired founder (at least from his pitch conviction) I met was a 60-year old in KL. 

 

> Certainly entrepreneurs don't see their role as being to serve investors - rather the other way around. Hence the very first thing that Brad Feld and David Cohen stressed when they spoke to Meng and I back in 2010 - the entrepreneurs are the only people that really create wealth in this picture. Everyone else is a servant or an enabler/block.

I deliberately use the word evolution as it really is a Darwinian process ... financial capital at minimum needs to be replaced. Sparks need chain-reactions for fission-fusion. And technology in the form of oxygen needs to be renewed. Wealth can be somewhat defined, whether productivity increase, new goods/services or just materialisation of intangibles (eg social graph + likes) ... if so the role of an accelerate is to bring the right combination together more efficiently than its peers. SV has an advantage, they tend to be more forward looking and capture ideas in earlier stage (needing less capital). Other places like NY or LA can mobilise huge talent pools and China simply dwarfs startups in sheer numbers. This is where govt policies and bringing it into mainstream (see China Indigenous Innovation policies) come in (my #1). 

Given that #3 culture (the social engineering side of entrepreneurship) is a nebulous influence, and #1 govt policies outsideside direct control of accelerators, perhaps the item most ammenable to private accelerators is accumulating unique resources, whether exclusive mentor networks (as Meng mentioned YC) or other hard to replicate capabilities.

[drllau]

Hugh wrote a 10-point perspective:

1. specialisation of accelerators - path dependency on local institutions
   
2. investor time-horizon & risk-appetite (related to general business conditions) dictate the type of accelerator
3. takes a village to raise a child, therefore spread the benefits whether monetary or vicarious
4. follow the money ... are accelerators aligned with cohort or sponsors (with potentially hidden agendas)
5. selection criteria - value-engineering or moonshot, and what risks are sponsors willing to take (team, tech, market, etc)
6. missing stages in cultural context of investment decision making - cf idea markets vs family patriarch vs stagegates
7. local variations or regional differences - cf Y-combo vs Startup Chile which lacked the initial entrepreneurial ecosystem
8. parameterisation (type of accelerator) based on objective data and prior frameworks
9. metrics (of success) are domain specific which also impacts on go-to-market (software lean-startup vs more deliberation in regulated industries)
10. diversification of entrepreneurship - not just ivy league frat boys but more democratisation of wealth-generation
    

I'm not totally convinced by the 3 horizons typology which to me seems more empirical (cf technical vs fundamental analysis in share markets). I conjecture with the right techno-economic knowledge, it is possible to predict certain trends and therefore attempt to time the markets (which is a high contributing factor to exit valuation). Specific individual trends like fashion or fads are hard to assess but the shift of complexity in 3D additive manufacturing, commoditisation of allied-health (innovation happens faster in non-regulated markets), the shift between existing household vs specialist outsourced functions, the millenial shift towards nomadic lifestyle (partly social disconnect, partly stimulation), all these trends can be extrapolated to derive the new markets which are only disruptive to incumbents who stick to the prevalent logic (cf Xerox and Japanese destop printers).

Lawrence

http://www.linkedin.com/in/drllau

[drllau]

When/where does a venture studio makes sense beyond colocation and mentoring? For a start, look at where the late stages make money and work backward. Conventional wisdom is based on 3 VC plays, adding value, sourcing better, and investing better. The issue with professional funds is that they are designed for a 10 year window which means harder to retain the human capital for HR, partnership channels, interoperatability, M&A. A venture studio which can be considered the management arm of a corporate VC fund (or multiple funds if sector agnostic) can retain these specialised functions. You might see this in places which have constrained markets (capital or otherwise) such as China where VC funds have Beijing office since they know the bottleneck is getting approval for IPO. Hence a venture studio can facilitate access to key resources needed for growth, or at least hold back the incumbants from crushing the upstarts (before buying out at a depressed valuation).

Much like the Hollywood model of distribution + small creative studios, the graduates of a venture studio should have higher quality, which means a better source of average deals (note outliers are completely different story). EntrepreneurFirst is not merely upstream co-founder team building, but using predictive analytics to stack the deck somewhat in their favor with the conclusion that generic business startup skills can be taught whereas hard technical edges are relatively scares. However, this assumes that they time the market right which, as owners of Newtons vs Phones know, can be a challenge.

Which comes down to timing and how to select investments for absolute alpha rather than minimising for beta (variability). As noted in article, art and science (or guts and brains) both play into picking. Value (and thus wealth) is created from improved productivity (eg Moore's Law which reduces unit energy per chip density as voltages drop), flattening distributions (eg internet consolidating many small markets), mobilising underutilised resources (eg crowd-sourcing like Wikipedia). As noted in Zero-to-One, its the market secret, identifying long-term trends to pick out weak signals in the talking head space, to be able to pick out proto-unicorns from the herd. Whilst a venture studio might not have any particular tech advantage, because they are often sector-based (cf Unearthed for extractive industries) they can ride several waves, (eg IoT, dematerialisation, AI) whereas a pure VC fund may be more fad driven which inherently prices the obvious candidates to above market value.

As you work backwards to earlier stages, the problems become manifest. An incubator doesn't have sufficient mass to justify the dedicated (human) resources for adding value. What I've seen in India is co-working spaces helping their clients organise their own events, effectively a form of self-learning as they teach each other necessary startup skills (marketing, paperwork, soft-skills, etc). Because there are limited spaces in accelerator cohorts, timing is harder so the obvious answer is to source better (hence the frequent appeals in mailbox asking to relocate to otherside of world), set up feeders (cf college football) or dedicated scouts. Being at the very early stages, weak signals are particularly hard to pick out whereas a late stage fund has a lot of social confirmation and often continental comparisons (cf China jumping into AI after recognising the lag against west). If a venture studio doesn't have artificial constraints like cohorts, and has recognised superior resources (mentors, corporate partnerships, testbeds, etc) over non-differentiated accelerators, with a sound investment thesis, then I can't see why they cannot outperform.

The majority of funds for startups are usually spent on people, either product engineering, DevOps or sales onboarding. Singapore has the chance to develop specialised venture studios around the various R&D resources in combination with the mix of markets from emerging-middle-developed in the region. The acquisition of flipkart shows how a big profile exit can energise the whole ecosystem. Who will be the first to prove the studio conjecture?

Lawrence

http://www.linkedin.com/in/drllau

[Rodrigo Martinez]

The “secret sauce” to building a successful accelerator is all here in this Harvard Business School case study: "NXTP Labs: An Innovative Accelerator Model” ( https://hbr.org/product/nxtp-labs-an-innovative-accelerator-model/815110-PDF-ENG )

Excerpt A: "Ariel [NXTP Labs Founder] had founded several companies and SOLD some of them shortly thereafter and made quite a bit of money.”

Excerpt B: "NXTP VALUED its network of mentors, who had DEEP expertise in starting, growing, scaling, and SELLING companies."

Key words in caps.

Cheers,

Rodrigo

https://rodrigomartinez.fyi.to/recommendations

--
You received this message because you are subscribed to the Google Groups "OpenFrog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openfrog+u...@googlegroups.com.
To post to this group, send email to open...@googlegroups.com.
Visit this group at https://groups.google.com/group/openfrog.
For more options, visit https://groups.google.com/d/optout.

[drllau]

Thanks Rod, NXTP was the first regional Latin American accelerator, as such whlist starting off a low base, it captured the low-hanging fruit. Hugh made note that you have to account for cultural and geographical specifics, a one-size doesn't fit all. Academics determine there are 3 types of capital: economic, social & cultural. Economics is easy to measure - multiple of money in but social capital (curated network of mentors, investors, founders, specialised services) is the village community to raise the startup child. Cultural capital is even more nebulous, the set of beliefs, norms and value levels that motivate and sustain entrepreneurship. Here I'd have to congratulate the SG govt for social engineering in encouraging more risk taking rather than seeking comfy govt sinceures. The reason why a lot of accelerators have been disappointing is that they assumed the presence of Silicon Valley social capital or trying to operate in a cultural desert (cf Thailand vs SG's initiatives). Just tossing buzzwords like LeanStartup methodology or design thinking without understanding the local context is setting yourself up to disappointment which is probably where a lot of post-exit founders have stumbled thinking to replicate their success without the necessary pre-requistites (also why accelerator space have become overcrowded)

To unsubscribe from this group and stop receiving emails from it, send an email to open...@googlegroups.com.