Monday, November 24, 2014

AlwaysOn Hosts OnMobile 2014 in Palo Alto

I am getting addicted to AlwaysOn events, and OnMobile 2014 this month was worth my time.  Tony Perkins' new Cuckoos Nest Club in Silicon Valley attracts big shots and naturally my presence completed the scene.  KPMG was a prominent sponsor and I was looking for their representatives on site.


Mr. Perkins' description of Millennial buying patterns got me thinking past their preference for sharing over privacy.  The Valley's fixation with Big Data isn't over because enterprises are still figuring out how to package it for resale.  I await the first startup that captures some network effect of Big Data aggregation.  We also have not seen the full maturation of AIs and machine learning that can magically generate video on demand, although a couple of startups in the AlwaysOn pipeline have some promising early solutions.  Fit all this stuff into mobile interfaces and watch the gross margins multiply into a bonanza.  

It's now a truism that apps count for more in the mobile UX than a mobile-optimized Web interface.  I heard that at countless conferences in 2013 and OnMobile drove the point home.  Big venture investors still chase early stage mobile companies but the biggest dumb money has barely arrived in the game.  I have sovereign wealth funds in mind when I think of dumb money.  They have no clue which technologies will win.  Startups with poor prospects may think that is good news if they get cash to burn.  A lot of corporate and strategic investors who are late to this game stand to lose a lot if they don't track sources like AlwaysOn.  

The fireside chat with Magisto was a classic showcase for how a first-mover advantage creates a network effect.  Magisto's video-editing AI already has peer credibility, a user base, and prominent investors.  A user base is still not the best traction metric in social media, especially with a freemium model like Magisto.  Sharing frequency should matter more because each sharing instance generates an ad impression.  Geographic segmentation of a user base matters in content creation for many reasons.  Not all alphabets read left-to-right like the Roman alphabet, for example.  The back-to-front visual narrative structure of Japanese manga is one geo-specific example that will pose a challenge to Western content creators.  

I won't repeat specifics from the startup pitches because some of what they shared is undoubtedly proprietary.  I will share my totally subjective general impressions.  Validation as "best app of the year" in the Google or Apple stores is certainly a worthy milestone.  A killer app would integrate AI auto-generated content with embedded data in its video and images, but I have not seen a good example ready for a market.  I can see using mobile apps for on-premise payments after service calls, but retail payments are integrating their POS proximity interfaces with links to apps for bank accounts and credit cards.  I expect service call support systems to simply adopt credit and debit card apps rather than reinvent the wheel.  

I'm seeing more startups in an emerging Silicon Valley archetype.  This archetype addresses only upper income market segments in developed countries who face lifestyle inconveniences.  They chase a lucrative but crowded segment whose disposable income enables price inelastic consumption for the time being.  The end of central bank intervention in asset markets will destroy that inelasticity.  More startups should at least try to address lower income segments and developing markets.  The world needs more than another lifestyle management app.  

The investor roundtable for B2B mobile predicted apps that could handle internal workflows, so the next hype cycle for enterprise workflows hasn't started until Gartner publishes something on workflow migration to mobile.  Real-time data flows will add value in fraud detection and micro-ad targeting but I doubt Gartner will get that specific.  Mobile ad attribution is complex and B2B trends are not making it any easier.  The roundtable people noted surges in mobile backend services and data rationalization for middleware.  Those two trends are IT-specific and I haven't yet seen how either one will help marketers understand ad attribution.  A rise in demand for APIs in those stacks should tell us whether marketers are using them to parse ad data.  

Another fireside chat with Auction.com showed how mobile is finally disrupting a real sector in the economy rather than layering another form of interaction on top of existing commerce portals.  Real estate sites are rapidly disintermediating traditional gatekeepers like banks and real estate agents.  Widely available online content de-risks deals by making due diligence easier.  I still think the potential for mobile to democratize access to real estate ownership is overestimated simply because REITs already exist and the economic barriers to owning them are very low.  

Enabling tech for B2B platforms warranted a separate panel.  Transaction sites obviously monetize from fees, but they can also display ads and sell their Big Data.  The startups that haven't figured that out are not getting the most out of their enabling tech.  I repressed a LOL when the panel said cryptocurrency enables a distributed trust environment for e-commerce.  The logic that a blockchain somehow enables a roving identity for shipping addresses is simplistic.  Clueless people underestimate the ability of determined cybercriminals to spoof or hijack a blockchain.  I'd rather go with layered standards as an industry-wide architecture for competing B2B tech solutions.  Personal identity is one layer and shipping address is a separate layer.  Different payment mechanisms (debit, credit, Apple Pay) get separate layers.  That's my end-run around blockchain weaknesses.  

I'll wrap up my analysis by suggesting a way ahead for mobile startups seeking attention at AlwaysOn and other Silicon Valley forums.  Startups appearing at pitchfests like OnMobile can build credibility by earning cloud computing awards.  There are too many to name here.  Ambitious startups should hit the Google search button and find ones they can earn.  They can do the same search for cloud sector metrics that bolster their business cases.  Cloud providers touching health care must be HIPAA compliant and mobile startups pitching workflows to the health care sector must also meet that standard.  If a blockchain approach is to have any place in a business model, it may be useful as a basis for identity management only if it is built once for reference and not expanded every time the identity holder makes a transaction.  Minimal coding means more rapid deployment, especially as iterative models replace SDLC waterfall models.  Entrepreneurs who talk more about their product than their business model will not connect with investors.  The path to monetization via a pricing model must be clear, and it must have an exit strategy.  Expected revenue growth models are becoming standardized thanks to Y Combinator and other accelerators, so startups can present more realistic estimates.  Anyone launching location-based services should first check with Gigaom Research to see who has done it before.  


I need to close with one big observation, and it segues from the 2D photo of the Olga Show host above.  Map APIs should enable 3D views.  The coming drone revolution means plotting air routes over short flights must account for altitude, terrain elevation, artificial obstructions, and other hazards to aviation.  Rotary-wing drones at higher altitudes won't generate as much lift in thinner air, so real-time information on air currents and other weather factors will be essential in next-gen 3D mapping.  The mobile startup that masters 3D mapping will own a big chunk of next-gen commerce.   You heard it hear first at Alfidi Capital.  I get my inspiration from AlwaysOn.