For AR/VR 2.0 to live, AR/VR 1.0 must die

The future of AR/VR could be bright, but only if it moves beyond where it is today. 2018 was the first of what look like two transitional years, with a potential shakeout in 2019 before an inflection point in late 2020. Let’s look at where we are today, where we’re heading tomorrow and some of the changes needed to get us there. (Note: There were earlier generations of AR/VR, but this discussion focuses on the post-2014 market.)

AR/VR Installed Base (including mobile AR)

Source: Digi-Capital AR/VR Analytics Platform

AR/VR 1.0 — where are we now?

AR/VR 1.0 kicked off in earnest when Facebook bought Oculus back in 2014. This inspired a generation of entrepreneurs, corporates and VCs to build early-stage AR/VR. Despite significant technical progress, even industry insiders admit this hasn’t produced a mass market yet.

Mobile AR delivered 2 percent higher revenue than we forecast for 2018 at over $3 billion globally, driven by app store revenues (primarily Pokémon GO), ad spend (e.g. from mobile AR features in messaging apps) and e-commerce sales (e.g. Houzz delivering 11x sales uplift). Mobile AR installed base (i.e. configured devices) grew more slowly than anticipated, to more than 850 million globally (note: this isn’t active users, which is a much lower number). As anticipated, there weren’t any standalone breakout mobile AR apps last year. Developers are still figuring out what does and doesn’t work for mobile AR.

Smartglasses had a mixed 2018, with Microsoft HoloLens winning a $480 million U.S. military contractMagic Leap launching more of a dev kit than a consumer product and other early smartglasses pioneers reported to be selling assets or furloughing staff. Smartglasses revenue (mainly hardware and enterprise solutions/services) was in the hundreds of millions of dollars, which together with mobile AR delivered total AR market revenue 3 percent lower than anticipated. So as in the last three years, AR revenue was broadly in line with Digi-Capital forecasts.

For VR, at the start of last year we didn’t anticipate phone makers largely abandoning mobile phone pre-order headset bundles (negatively impacting mobile VR sales/installed base significantly), and got timing wrong for Oculus Quest launching in holiday 2018 (announced late last year as Spring 2019). The mid-year launch of Oculus Go and our accurate forecast of Sony PSVR sales helped, but together with attrition rates, the VR market was down year-on-year for 2018 in terms of unit sales, installed bases and revenue at less than $3 billion (rather than the modest growth we forecast).

AR/VR 2.0 — where are we going?

AR (mobile AR, smartglasses) could top two and a half billion installed base and $70 billion to $75 billion revenue by 2023. VR (mobile, standalone, console, PC) might deliver more than 30 million installed base and $10 billion to $15 billion revenue in the same time frame. That’s a pretty big difference, so let’s dig into the data to understand why.

AR/VR platform revenue

Source: Digi-Capital AR/VR Analytics Platform

Mobile AR

While mobile AR revenue outperformed slightly last year, underlying data at the platform level has guided us to downgrade mobile AR’s installed base long term. Where Apple and Facebook control their platforms (ARKit, Spark AR), Google doesn’t — it had to rely on Android phone maker partnerships to grow ARCore configured devices from 100 million to 250 million last year.

That’s still a big number, but the growth curve it implies now means that our AR/VR Analytics Platform forecasts ARCore’s installed base trailing Apple/Facebook until 2021. So while ARKit and Spark AR growth paths remain on track with previous forecasts, a slower growth path for ARCore means there could be a total mobile AR market installed base just over two and a half billion globally by 2023. Again, still a big number, just not as big as originally anticipated. 

Mobile AR business model revenue

Source: Digi-Capital AR/VR Analytics Platform

E-commerce across 10 major categories (from cars to clothing to toys) looks set to be mobile AR’s largest revenue stream, which together with ad spend across 11 major advertiser categories (from retail to CPG to travel) could deliver three quarters of mobile AR revenue long term. 

AR e-commerce sales

Source: Digi-Capital AR/VR Analytics Platform

Mobile AR app store revenues (in-app-purchase/premium) remain dominated by games (primarily Pokémon GO) today, but mobile AR’s installed base and increasing adoption as a feature in mainstream apps could see non-games categories take more than half the mobile AR app store revenues by 2023. As with the overall mobile market, it could remain difficult for standalone mobile AR apps to rise to the top of app stores. Mobile AR growth could have more to do with mobile AR features in ubiquitous apps than new standalone apps.

Mobile AR app store categories revenue

Source: Digi-Capital AR/VR Analytics Platform

Smartglasses

Smartglasses have to deliver on five major challenges before they can become mass market consumer devices: (1) hero device (i.e. an Apple quality device, whether made by Apple or someone else), (2) all-day battery life, (3) mobile connectivity, (4) app ecosystem and (5) price. Together, these are non-trivial problems, and could see smartglasses remain mainly enterprise focused through the middle of 2020. Smartglasses device sales could stay in the hundreds of thousands of units globally this year.

Smartglasses business model revenue

Source: Digi-Capital AR/VR Analytics Platform

If Apple launches iPhone-tethered smartglasses in late 2020 (as we’ve forecast since 2016), the AR/VR market could finally see its inflection point. Nonetheless, long-term smartglasses revenues could remain dominated by hardware and enterprise (ex-hardware) revenues through 2023. The mass market for consumer smartglasses still looks a long way off, even with Apple’s entry. 

Smartglasses enterprise revenue

Source: Digi-Capital AR/VR Analytics Platform

Smartglasses enterprise pilot projects and full-scale rollouts have been symptomatic of an early-stage tech platform, but real-world productivity results are now being delivered with companies like Lockheed Martin reducing satellite building activities by more than 50 percent using HoloLens/Scope AR. When smartphone-tethered smartglasses reduce system costs and expand the range of applications, “bring your own device” could see smartglasses enterprise revenues kickstart an inflection point in 2021 across manufacturing/resources, TMT, government (including military), retail, construction/real estate, healthcare, education, transportation, financial services and utilities industries.

VR

VR could return to modest growth this year, and remain dominated by hardware and games revenues. The second generation of premium standalone VR headsets (not those launching this year) could become a catalyst in the 2020/2021 time frame. For this to happen, they will need to deliver greater performance and better content at lower prices. Hopefully by that time we could also see VR platform holders simplify their product ranges from their current platform fragmentation (taking a page out of Steve Jobs’ 1997 playbook).

VR business model revenue

Source: Digi-Capital AR/VR Analytics Platform

VR revenue comes primarily from entertainment, and is driven by premium/standalone VR more than mobile/standalone VR due to installed bases and unit economics. Games software revenue could dominate long term, followed by hardware, enterprise (ex-hardware), video and location-based entertainment revenue streams. Due to VR platform holders’ gamer focus, they face the same challenges as Sony and Microsoft when they tried to diversify games console revenue streams beyond games (with mixed results).

AR/VR countries revenue

Source: Digi-Capital AR/VR Analytics Platform

Asia is set to dominate AR/VR for the next five years, driving more revenue than North America and Europe combined by 2023. China’s commitment to the market is a standout, and it could remain the largest single market for AR/VR long term.

So what’s needed to go from AR/VR 1.0 to 2.0?

A lot of things might need to change to take us from AR/VR 1.0 to 2.0:

High friction to low friction: A large part of AR/VR 1.0 remains high friction in terms of installation, UX and UI. In many ways the market today looks like the MP3 player market before Steve Jobs launched the iPod (keep that analogy in mind). AR/VR 2.0’s lower friction is in the works, but what’s needed here are Apple smartglasses (whether they call them iGlasses or something else), the second generation of premium standalone VR (that comes after Oculus Quest and HTC Vive Focus) and mobile AR developers innovating beyond the lessons learned from Niantic, Houzz and others.

Experiences to use cases: There have been many “experiences” during AR/VR 1.0, with visually stunning apps not delivering meaningful UX. An AR/VR dragon or portal is impressive the first time you see it, but gets old pretty quickly. The next stage of AR/VR must deliver against critical use cases, with features in critical apps that we use all day, every day.

Standalone to features: The industry has largely focused on standalone apps to date, but major features in apps we use every day could see higher usage and prove more commercially successful. Navigation (Google Maps), e-commerce (AmazonWalmartAlibaba) and messaging (Facebook Spark ARSnapchat Lens Studio) are beginning to show how this might happen.

Expensive to good value: Early AR/VR has ranged from $3,000 HoloLens to $200 Oculus Go to free mobile AR. But competing platforms often deliver more for less outside of specific use cases, particularly where we already own them, as with mobile. AR/VR 2.0 needs to become a great value because of what it delivers to users regardless of price point.

Point solutions to ecosystem: Many early AR/VR apps have been entertainment (games, video) or standalone point solutions to specific problems. As we’ve discussed before, AR/VR needs its own reality ecosystem to scale.

Low ROI to high ROI: For consumers, this means apps that give back more than just a “wow,” and for enterprises, applications that deliver real return on investment. This is beginning to happen in enterprise with companies like Lockheed Martin and Bell.

Pilots to production: Enterprise AR/VR 1.0 has seen many pilot projects, but relatively few full production rollouts. This is beginning to change, with companies like Walmart (with STRIVR) beginning to move into full production.

Inside baseball to brands: The AR/VR industry is still debating the merits of using AR, VR, MR, XR or spatial computing to describe itself, as well as spending a lot of time focused on internal plumbing across the stack. But consumers and enterprises outside of early adopters don’t care. They buy brands that deliver against critical use cases rather than just tech, which requires a clearer focus on users and how to market to them to succeed.

Fragmentation to dominance: AR/VR 1.0 remains fragmented across both hardware and software, despite its early stage and relatively small user bases. The industry now appears to have made up its mind on which platforms matter, so natural selection could thin the herd to a few dominant players in each part of the market.

Blue Sky to data driven: Many AR/VR 1.0 companies have been coy about their numbers, with independent data sources hard to come by in the early market. Developments like Digi-Capital’s AR/VR Analytics Platform have made it difficult to hide, with hard data/analytics now available to answer granular questions about roadmaps, country rollouts, investments, valuations and more.

VC-funded to cockroach/money making: Well-funded pioneers began to exit the market last year, with 2019 potentially seeing a major shakeout of companies that aren’t at least breaking even. The U.S. AR/VR investment market began to reverse its decline in Q4 2018 (even as Chinese investment accelerated), but making money and “cockroaching” burn rate could be more important than VC money in AR/VR 2.0.

Everyone else to Apple: If Apple launches smartphone-tethered smartglasses in late 2020, AR/VR 2.0 could have its “iPod moment,” where a major new form factor introduces the starting point for a long-term mass consumer market. It’s worth noting that this might not be the industry’s “iPhone moment,” as even with this catalyst we aren’t looking at a mass consumer market in the next five years.

Denial to acceptance: 2019 isn’t the “Year of AR/VR,” and Mark Zuckerberg’s “1 billion people in VR” might never happen. Mark’s come to terms with it, so hopefully a sense of cautious optimism could prevail during the next stage of the market.

What about AR/VR 3.0?

Even though we’re looking at a potential $80 billion to $90 billion AR/VR market by 2023, AR/VR 2.0 won’t be the finished article. That could take a lightweight pair of standalone smartglasses, capable of replacing your iPhone at the same price. There are formidable technical and content challenges to reach that vision of AR/VR 3.0, and there’s AR/VR 2.0 to navigate first.

It’s going to be an exciting time, and we can’t wait to see what comes next.

Are rightsholders ready for public domain day?

On January 1, 2019, the New Year will ring in untold numbers of additions to the public domain in the U.S., including hundreds and maybe thousands of works with at least a small public reputation. This, of course, is due to the expiration of the terms of their copyrights, some of which have been extended multiple times since the 1960s. 

This is a good thing from many perspectives, including that of authors, publishers, museum curators, teachers, old-book readers and music and film buffs. It possibly may be a slightly bad thing for a few people — primarily certain estates representing long-dead authors and other creators.

What’s a “term” in the context of copyright?

The duration, or term, of U.S. copyright is set by Congress, and has gradually crept up over time from the original 14 years (plus 14 more if the author was still alive and renewed the copyright) — in Thomas Jefferson’s time — to a whopping “life of the author plus 70 years,” as set by the 1998 “Copyright Term Extension Act” (CTEA, which extended it from life plus 50).

For works first published between 1909 and 1978, the maximum term was finally set by Congress at 95 years (assuming the author complied with a whole lot of rules, alluded to below).  And for post-1978 works, in instances where the author/creator is not a human being (such as a business commissioning a “work made for hire” under rules developed in the case-law) or the work was published under a pseudonym for an unknown person, the term can be as long as 120 years! The copyright in a work, duly registered at the time that registration was required (pre-1978), may never have been renewed, and so its protection may have quietly lapsed some time ago; for many more obscure works, it’s hard to know.

Fun fact: This Copyright Term Extension Act is also known as the Sonny Bono Copyright Term Extension Act. Congress named it in memory of the composer of “I’ve Got You, Babe,” who, as a member of Congress from Southern California, was among the authors of the bill; he unfortunately happened to die while it was being worked on in committee. Prior to 1978, the term of U.S. copyrights was determined by fixed terms of years, subject to publication, registration and notice requirements. Here are more details on that.

How do works pass into the public domain?

Currently, works pass into the public domain according to a complex schedule, combining (sometimes awkwardly) the rules of various laws implemented over the past century.

Bear in mind, however, that many works have passed (or “fallen” or “lapsed,” as the older phrases had it) into the public domain in the U.S. for reasons other than term expiry, even during the 20 years of the CTEA extension. According to the law in effect prior to 1978, if the work was published but never registered in the U.S. Copyright Office, it did not receive protection under copyright law; a work might also not be protected by U.S. copyright law if it lacked proper notice — the © symbol and the proper wording — or if the work’s registration was not renewed after its first 28-year term expired. Or if, as a work of the federal government, it never enjoyed copyright protection in the first place.

Qui Bono? (get it?)

As it turns out, it is not just re-publishers of “classic” texts, such as Dover Thrift Editions, which benefit when new works become available. Textbook and educational publishers frequently re-use old short stories and essays in larger collections, and a work of marginal utility might become more attractive as a potential addition to these collections once the cost of clearing the rights is reduced.

For example, a few years ago a 1922 story by F. Scott Fitzgerald, “The Curious Case of Benjamin Button,” (whose U.S. copyright had lapsed) was adapted into a feature film. To me, the lesson to be gleaned is that many works of the early 20th century still appear to bear some cultural cachet (or at least continuing value to society) — such that more no-cost access to these works (by their passing from copyright protection to the public domain) should have the overall effect of helping them find new audiences.

Note: Bear in mind, all of these examples are simply illustrative — without a full and careful copyright search, it is difficult to be certain of the copyright status of almost any work. On that, more below.

New works coming into the U.S. public domain also will have the effect of giving researchers new texts to run Text and Data Mining (TDM) algorithms across. It also may add to the richness of film and cultural studies.

Mark Twain proves this isn’t so easy

Unfortunately, determining when a work has in fact “fallen” into the public domain due to the term of its copyright having expired is not always as simple as one might hope.

For example, one might think that everything ever laid down by the pen of Mark Twain (S.L. Clemens, d. 1910) would be in the public domain by now. But, since he left a treasure trove of unpublished works, their copyright protection has extended for many years after his death, because, under pre-1978 law, those works’ copyright protection would not start until the works were published. The distinction between published and unpublished works has been discarded under post-1978 law, but won’t be fully effective for another 30 years. So, some items in the microfilm edition of Twain’s letters and manuscripts (their first publication) are still considered to be under copyright. He’s also enjoyed considerable success recently with the full and final publication of his autobiography.

Twain, a student of intellectual property, steadfastly argued for a perpetual copyright, but he came to realize that this was not permitted under the copyright clause of the U.S. Constitution, which refers to “securing [protection] for limited times.” But, in an age when copyright only protected works for which registrations had been obtained, he did point out that most books wouldn’t be affected by a longer term at all — for the vast bulk of them had no commercial life remaining to them a very few years after their initial publication:

One author per year produces a book which can outlive the forty-two-year limit; that’s all. This nation can’t produce two authors a year that can do it; the thing is demonstrably impossible. All that the limited copyright can do is to take the bread out of the mouths of the children of that one author per year.

I made an estimate some years ago, when I appeared before a committee of the House of Lords, that we had published in this country since the Declaration of Independence 220,000 books. They have all gone. They had all perished before they were ten years old. It is only one book in 1000 that can outlive the forty-two-year limit. Therefore, why put a limit at all? You might as well limit the family to twenty-two children.

– S.L. Clemens, in testimony to Congress, concerning proposed copyright legislation (1906)

“Forever minus a day,” another idea which has been occasionally bruited about (particularly by Congressman Bono and his widow, who was later elected seven times in her own right to Congress), would not constitute much of an effective limit, and so would, I believe, violate the Constitutional limitation; 95 years (an estimated average of the “Life plus 70” term) seems closer to a natural lifespan for a copyright — to me at least. If you and your heirs somehow can’t get the commercial value out of your work before nearly a century is out, I think there’s a takeaway lesson there.

On the other hand…

… some works do have cultural lifespans exceeding the term of copyright. The estates of certain literary, film and musical creators may stand to lose when the copyright in some of the works in their respective repertories lose copyright protection due to the lapse of their terms. For some examples of works entering the public domain on January 1, 2019, that may still have financial value to the author/creator’s heirs: Hemingway’s “Three Stories and 10 Poems” was first published in 1923; it was also the year of release for “Safety Last!” a silent film from Hal Roach Studios, starring Harold Lloyd, which many people remember. The same year saw the first publication (of the sheet music) for “Who’s Sorry Now?” which was a hit recording for Connie Francis in 1958.

But, on balance, “Nothing gold can stay,” as Robert Frost observed in a poem slated — I’m pretty sure — to enter the public domain on January 1st.* The reading, listening, and viewing public should expect to be the main beneficiary of these works entering the public domain. Indeed, 95 years is a good run for the commercial exploitation of a work. Now it’s everybody else’s turn to benefit.

*If it hasn’t already. Copyright searches, on the detail level, can be quite difficult and time-consuming. See: https://www.copyright.gov/rrc/. For any proposed commercial republication, it is certainly the course of wisdom to consult with an attorney and have a full copyright search done.

Venture capital, global expansion, blockchain and drones characterize African tech in 2018

2018 saw Africa’s tech sector become more dynamic and international. VC firms on the continent multiplied. There were numerous investment rounds. And startups pursued acquisitions and global expansion. Here’s a snapshot of the news that shaped African tech over the last year.  

Surge in VC funds

A notable 2018 trend was Africa’s VC landscape becoming more African, with an increasing number of investment funds headquartered on the continent and run by locals, according to Crunchbase data released in this TechCrunch exclusive.

Drawing on its database and primary source research, Crunchbase identified 51 viable Africa-focused VC funds globally with at least 7-10 investments in African startups from seed to series stage.

Of the 51 funds, 22 (or 43 percent) were headquartered in Africa and managed by Africans. Of those 22, nine (or 41 percent) were formed since 2016 and nine were Nigerian.

Four of the nine Nigeria-based funds were formed within the last year: Microtraction, Neon Ventures, Beta.Ventures and CcHub’s Growth Capital fund.

The Crunchbase study also tracked more Africans in top positions at outside funds and the rise of homegrown corporate venture arms.

One of those entities with a corporate venture arm, Naspers, announced a $100 million fund named Naspers Foundry to invest in South African tech startups. This was part of a $300 million (4.6 billion Rand) commitment by the South African media and investment company to support South Africa’s tech sector overall, as reported here at TechCrunch.

Another DFI came on the scene when France announced a $76 million African startup fund administered by the French Development Agency, AFD. TechCrunch got the skinny on how it will work here.

Investment and expansion

If African VC investment headlines were scarce a decade ago, in 2018 we became overwhelmed with them. This was largely a result of several recently closed Africa funds — TLcom’s $40 million, Partech’s $70 million, TPG’s 2 billion — beginning to deploy that capital.

In March, Nigerian consumer data analytics firm Terragon raised $5 million from TLcom. Kenyan business enterprise software company Africa’s Talking raised $8.6 million in a round led by IFC.

Investment startup Piggybank.ng closed $1.1 million in seed funding and announced a new product — Smart Target, for traditional savings groups. Trucking Logistics company Kobo360 raised two rounds, for a total of $7.2 million. Kenya-based agtech supply chain startup Twiga Foods raised $10 million. B2B retail supply chain Sokowatch closed a $2 million seed round led by 4DX ventures.

White-label lending startup Mines.io secured a $13 million Series A round. South African SME payment venture Yoco raised $16 million. Paga Payments added $10 million in fresh funding.

And then there were the three huge raises of the year. Kenyan digital payment company Cellulant hauled in $37.5 million in a Series C round led by TPG Growth. South African lending startup Jumo raised $52 million led by Goldman Sachs. And just this month, The Carlyle Group invested $40 million in Africa-focused online travel site Wakanow.com.  

Acquisitions and expansion

In 2018, African tech demonstrated it can travel, as several digital companies expanded on the continent and abroad. In May, MallforAfrica and DHL launched MarketPlaceAfrica.com, a global e-commerce site for select African artisans to sell wares to buyers in any of DHL’s 220 delivery countries.

Paga announced plans to expand in Africa and internationally, with an eye on Ethiopia, Mexico and the Philippines, CEO Tayo Oviosu told TechCrunch. Kobo360 is moving into in new markets — Ghana, Togo and Cote D’Ivoire.

On the back of its $52 million round, Jumo said it would expand in Asia and started by opening an office in Singapore.

On the acquisition front, Terragon bought Asian mobile marketing company Bizense in a cash and stock deal. The company is exploring greater growth opportunities in Latin America and Southeast Asia, CEO Elo Umeh told TechCrunch.

TPG Growth acquired a majority stake (of an undisclosed value) in Africa entertainment content company TRACE. After previous investments, Naspers acquired  96 percent of Southern African e-commerce venture Takealot.

And in December, California-based Emergent Technology Holdings acquired Ghanaian fintech payment company InterpayAfrica.

Partnerships

Collaboration between local tech firms and big global names continued in 2018. Liquid Telecom and Microsoft continued their partnership to offer connectivity cloud services such as Microsoft’s Azure, Dynamics 365 and Office 365 to select startups and hubs. This is part of Liquid Telecom’s strategy to go long on Africa’s startups as its future clients and the continent’s next big companies.

Facebook teamed up with Nigerian tech hub CcHub to launch its NG_Hub high-tech incubator.

Blockchain

As crypto fever gripped many leading economies in 2018, Africa was shaping its own blockchain narrative — one more grounded in utility than speculation. 500 Startups-backed SureRemit launched a crypto token product aimed at disrupting Africa’s multi-billion-dollar remittance market and raised $7 million in an ICO. South African payments venture Wala and solar energy startup Sun Exchange also had ICOs.

For blockchain as a platform, agtech startups Twiga Foods and Hello Tractor partnered with IBM Research to use the digital ledger tech to advance small-scale farmers and agriculture on the continent.

Ride-hail boda bodas

Ride-hail tech expanded into the continent’s frequently used motorcycle taxi market. Uber entered the three-wheeled tuk tuk moto taxi market in Tanzania in March and Uber and Taxify launched motorcycle passenger services in East Africa, including Kenya and Uganda.

Fails

Last year saw Y Combinator-backed VOD startup Afrostream shutter. In February 2018, Nigerian e-commerce startup Konga — backed by VC — was sold in a distressed acquisition. There were high expectations for Konga and its much-liked founder Sim Shagaya. I made the case that Konga’s acquisition was one of Africa’s first big startup fails that flew under the radar.

Drones

TechCrunch did a deep dive into Africa’s drone scene, talking to several experts and looking at emerging use cases across delivery services, agtech and surveying. On the regulatory side, several countries — Rwanda, Tanzania, South Africa, Zambia and Malawi — are doing some interesting things around regulation and creating drone-testing corridors for global players.

TechCrunch and Africa

In 2018 TechCrunch did more with Africa than any previous year. In addition to more content, there was a market engagement trip to Ghana and Nigeria, with meet-and-greets at Impact Hub, MEST Accra and Lagos, and CcHub.

TechCrunch also had its first Africa panel on Disrupt SF’s main stage, an Africa session at Disrupt Berlin and held the second Startup Battlefield Africa in December in Nigeria.

Fifteen startups competed in Lagos in front of a Pan-African and global crowd. South African virtual banking startup Bettr was runner-up. Ultra-affordable ultrasound startup M-Scan from Uganda was the winner.

More Africa-related stories @TechCrunch

African tech around the ‘net  

Why your startup shouldn’t rush to $1 million in revenue

There is a prevailing belief that the magic formula for early-stage tech startups hinges on how quickly they achieve $1 million in annual recurring revenue (ARR). Investors in SaaS companies, in particular, are very guilty of pushing this or its equally loaded corollary, “When will you sign your first six-figure deal?”

But in the rush toward these numbers, too many startups lose sight of their primary intent: These metrics are supposed to be an indicator of product/market fit. We’ve seen companies reach $1 million in ARR in less than a year, yet not have enough market momentum to get their next million easily. We’ve seen early-stage companies so concerned about getting those first sales, they don’t validate the market and if they’re building the right product. We’ve also watched a focus on new logos make companies forget about keeping existing customers happy, introducing unexpectedly high churn — something startups can’t afford.

Those first customers and that first million are supposed to be the bedrock on which the rest of the business grows. Founders must constantly ask what they’re learning about their market, product and go-to-market approach — in that order! — so the business becomes a flywheel.

Revenue is a lagging indicator of sales success, so must likewise be prioritized accordingly. That’s not to say revenue isn’t vitally important and that there isn’t a great deal of urgency to it, but focusing on it too much too early can mask big problems that will hurt startups later when the stakes are higher.

Here are a few lessons we’ve learned by watching our early-stage companies go through this crucial phase. Every early-stage company needs to do them well.

Customer and market discovery is job No. 1

We talk about product and knowing customers a lot, but that is insufficient. Startups must understand the market, as well. How do customers do this today? Is there urgency around the problem? What is the community saying? An early investor in PagerDuty went onto Reddit and Quora and just looked at who people were talking about. It made his decision easy.

To be really successful, it is as important to understand market dynamics as it is to deliver a great product. This also helps zero in on all the aspects of your ideal customer profile; it needs to be more specific than you think! This also then helps qualify customers for future sales.

Elevate Security stood out in their super-crowded security space because they carved out a unique position around people-powered security. They used their early sales process to carefully qualify who would help them best develop their products. Their first product got shout-outs on social media from users who loved it — a rare occurrence in security — and were indicators they had found good initial customers and were creating something unique.

Build a product that sells itself

You’ll always find smart people saying, “I love what you’re doing.” Some things are so broken even a mediocre improvement is worth a change. But this is why revenue can be a false indicator for scalable success: Founders find enough early adopters to get that first million, which leads them to believe the product is enough. The company starts chasing more revenue, not investing in a product-based growth engine. If sales keeps hitting their numbers, everyone believes things are fine. Until they’re not. And then it’s usually a really heavy lift, with 6-12 months of product, sales or team upgrades.

What startup doesn’t want a growth curve like this? Zoom had triple-digit growth for the last four years in a crowded, mature video conferencing category. Janine Pelosi, Zoom’s head of marketing, said the reason they were so successful before and after she arrived was they have a great product. It’s reliable, easy to use, and the founder, Eric Yuan, was selling it every day. Yuan knew the market really well coming out of Webex, and always touching customers meant he could adjust company strategy accordingly. Zoom embodied the real magic formula: know your market + build great product.

Pay attention to customer engagement and delight

Customer satisfaction is simple: It comes from the perception that people get value from their purchase; it’s much less about how much they paid. It’s also always cheaper to make an existing customer happy than it is to acquire a new one, so make sure even in the early days that you’re investing in making current customers happy advocates.

Aquabyte uses computer vision to identify sea lice in the $160 billion aquafarming market. When they showed customers FreckleID (think racial recognition for fish) to uniquely identify fish in a pen of 200,000, fish farmers loved the idea. The price they were willing to pay was 3x what the CEO thought possible. They’re likewise investing heavily in making sure their initial customer is successful with the product and are delighting them in unexpected ways (handwritten holiday cards). They have more prospects in their pipeline than they have capacity, which means they don’t need to expand sales to grow revenue fast.

Your startup may have the coolest tech, be in the biggest market and have the smartest team. No matter what your board says, remember revenue is NOT the primary indicator; it is simply an indicator. To become a breakout success, you need to read the tea leaves of all aspects of your market and build a product and customer experience that is truly superior.

Cyber breaches abound in 2019

News of high-profile cyber breaches has been uncharacteristically subdued in recent quarters. However, we recently learned that Marriott International/Starwood was the victim of the multi-year theft of personal information on up to 500 million customers — rivaled only by hacks against Yahoo in 2013 and 2014.

Is this a harbinger of a worse hacking landscape in 2019?

The answer is unequivocally yes. No question, cyber breaches have been a gigantic thorn in the global economy for years. But expect them to be even more rampant in the new year as chronically improving malware will be deployed more aggressively on more fronts.

In addition, as companies increasingly pursue digitization to drive efficiency, reduce costs and build data-driven businesses, they simultaneously move into the “target zone” of cyber attacks. As the digital economy expands, the threat landscape naturally follows suit. Compounding the situation is the use of machine learning and AI as hackers and other bad actors look to scale their bad behavior.

Look for AI-driven chatbots to go rogue, a substantial increase in crimeware-as-a-service, acceleration of the weaponization of data, a resurgence in ransomware and a significant increase in nation-stage cyberattacks. Also on a growth track is so-called cryptojacking — a quiet, more insidious avenue of profit that relies on invasive methods of initial access and drive-by scripts on websites to steal resources from unsuspecting victims.

Then, too, we will also see a substantial increase in software subversion, including the specific targeting of developers for attack and the likely proliferation of software update supply chain attacks.

Here is a mini dive into the top pending threats:

The emergence of AI-driven chatbots. In the new year, cybercriminals and black hat hackers will create malicious chatbots that try to socially engineer victims into clicking links, downloading files or sharing private information. A hijacked chatbot could easily misdirect victims to nefarious links rather than legitimate ones. Attackers are also likely to leverage web application flaws in legitimate websites to insert a malicious chatbot into a site that doesn’t have one.

Attacks on cities with crimeware-as-a-service, a new component of the underground economy. Adversaries will leverage new tools that among other things attack data integrity, disabling computers to the point of requiring mandatory hardware replacements. Terrorist-related groups will be the likely culprits.

A significant increase in nation-state attacks. Russia has been a leader in using targeted cyberactions as part of larger objectives. Earlier this year, for example, the FBI disclosed that Sofacy group, a Russian persistent threat actor, infected more than 500,000 home office routers and network attached to storage devices worldwide to remote control them. Look for other nation-states to follow the same sort of playbook, helped by billions of poorly secured IoT devices.

The growing weaponization of data. Already a huge problem, it is certain to worsen, notwithstanding efforts among some technology giants to enhance user security and privacy. Balancing the negatives with the positives, tens of millions of comprised web users have begun to seriously question how much they really benefit from the internet.

Consider, for example, Facebook, which has made no secret of using personal data and “private” correspondence to annually generate billions of dollars in profits. Users willingly “like” interests and brands, volunteering personal information. This enables Facebook to provide a more complete image of its user base — a gold mine for advertisers.

Much worse, Facebook earlier this year tried to manipulate user moods through an “emotional contagion” experiment. This pitted users against their peers to influence their emotions, i.e. the weaponization of data.

A resurgence in ransomware. Ransomware exploded onto the scene in 2017 following the WannaCry outbreak and a series of successful follow-up ransomware attacks targeting high-profile victims. According to the FBI, total ransomware payments in the U.S. have in some years exceeded $1 billion. There were scant high-profile ransomware victims in recent months, but the problem is highly likely to bounce back strongly in 2019. Ransomware attacks come in waves, and the next one is due.

Increased subversion of software development processes and attacks on software update supply chains. Regarding software development, malware has already been detected in select open-source software libraries. Meanwhile, software update supply chain attacks violate software vendor update packages. When customers download and install updates, they unwittingly introduce malware into their system. In 2017, there was an average of one attack every month, compared to virtually none in 2016, according to Symantec. The trend continued in 2018 and will become worse next year.

More cyber attacks on satellites. In June, Symantec reported that an unnamed group had successfully targeted the satellite communications of Southeast Asia telecom companies involved in geospatial mapping and imaging. Symantec also reported attacks originating in China last year on a defense contractor’s satellite.

Separately, we learned in August at the annual Black Hat information security conference that the satellite communications used by ships, planes and the military to connect to the internet are vulnerable to hackers. In the worst-case scenario, the research said, hackers could carry out “cyber-physical attacks” that could turn satellite antennas into weapons that essentially operate like microwave ovens.

Fortunately, the cyber outlook for 2019 is not altogether grim.

On the cybersecurity side, a growing number of experts believe that multi-factor authentication will become the standard for all online businesses, abandoning password-only access. In addition, a number of states are expected to adopt some version of Europe’s strict General Data Protection Legislation. California, for one, has already passed legislation that will make it easier for consumers to sue companies after a data breach, starting in 2020.

The upshot is that individuals, businesses and government entities need to do everything possible to improve the state of their cybersecurity. They cannot eliminate breaches, but they can avert some and improve the chances of mitigating them.

Cyber breaches abound in 2019

News of high-profile cyber breaches has been uncharacteristically subdued in recent quarters. However, we recently learned that Marriott International/Starwood was the victim of the multi-year theft of personal information on up to 500 million customers — rivaled only by hacks against Yahoo in 2013 and 2014.

Is this a harbinger of a worse hacking landscape in 2019?

The answer is unequivocally yes. No question, cyber breaches have been a gigantic thorn in the global economy for years. But expect them to be even more rampant in the new year as chronically improving malware will be deployed more aggressively on more fronts.

In addition, as companies increasingly pursue digitization to drive efficiency, reduce costs and build data-driven businesses, they simultaneously move into the “target zone” of cyber attacks. As the digital economy expands, the threat landscape naturally follows suit. Compounding the situation is the use of machine learning and AI as hackers and other bad actors look to scale their bad behavior.

Look for AI-driven chatbots to go rogue, a substantial increase in crimeware-as-a-service, acceleration of the weaponization of data, a resurgence in ransomware and a significant increase in nation-stage cyberattacks. Also on a growth track is so-called cryptojacking — a quiet, more insidious avenue of profit that relies on invasive methods of initial access and drive-by scripts on websites to steal resources from unsuspecting victims.

Then, too, we will also see a substantial increase in software subversion, including the specific targeting of developers for attack and the likely proliferation of software update supply chain attacks.

Here is a mini dive into the top pending threats:

The emergence of AI-driven chatbots. In the new year, cybercriminals and black hat hackers will create malicious chatbots that try to socially engineer victims into clicking links, downloading files or sharing private information. A hijacked chatbot could easily misdirect victims to nefarious links rather than legitimate ones. Attackers are also likely to leverage web application flaws in legitimate websites to insert a malicious chatbot into a site that doesn’t have one.

Attacks on cities with crimeware-as-a-service, a new component of the underground economy. Adversaries will leverage new tools that among other things attack data integrity, disabling computers to the point of requiring mandatory hardware replacements. Terrorist-related groups will be the likely culprits.

A significant increase in nation-state attacks. Russia has been a leader in using targeted cyberactions as part of larger objectives. Earlier this year, for example, the FBI disclosed that Sofacy group, a Russian persistent threat actor, infected more than 500,000 home office routers and network attached to storage devices worldwide to remote control them. Look for other nation-states to follow the same sort of playbook, helped by billions of poorly secured IoT devices.

The growing weaponization of data. Already a huge problem, it is certain to worsen, notwithstanding efforts among some technology giants to enhance user security and privacy. Balancing the negatives with the positives, tens of millions of comprised web users have begun to seriously question how much they really benefit from the internet.

Consider, for example, Facebook, which has made no secret of using personal data and “private” correspondence to annually generate billions of dollars in profits. Users willingly “like” interests and brands, volunteering personal information. This enables Facebook to provide a more complete image of its user base — a gold mine for advertisers.

Much worse, Facebook earlier this year tried to manipulate user moods through an “emotional contagion” experiment. This pitted users against their peers to influence their emotions, i.e. the weaponization of data.

A resurgence in ransomware. Ransomware exploded onto the scene in 2017 following the WannaCry outbreak and a series of successful follow-up ransomware attacks targeting high-profile victims. According to the FBI, total ransomware payments in the U.S. have in some years exceeded $1 billion. There were scant high-profile ransomware victims in recent months, but the problem is highly likely to bounce back strongly in 2019. Ransomware attacks come in waves, and the next one is due.

Increased subversion of software development processes and attacks on software update supply chains. Regarding software development, malware has already been detected in select open-source software libraries. Meanwhile, software update supply chain attacks violate software vendor update packages. When customers download and install updates, they unwittingly introduce malware into their system. In 2017, there was an average of one attack every month, compared to virtually none in 2016, according to Symantec. The trend continued in 2018 and will become worse next year.

More cyber attacks on satellites. In June, Symantec reported that an unnamed group had successfully targeted the satellite communications of Southeast Asia telecom companies involved in geospatial mapping and imaging. Symantec also reported attacks originating in China last year on a defense contractor’s satellite.

Separately, we learned in August at the annual Black Hat information security conference that the satellite communications used by ships, planes and the military to connect to the internet are vulnerable to hackers. In the worst-case scenario, the research said, hackers could carry out “cyber-physical attacks” that could turn satellite antennas into weapons that essentially operate like microwave ovens.

Fortunately, the cyber outlook for 2019 is not altogether grim.

On the cybersecurity side, a growing number of experts believe that multi-factor authentication will become the standard for all online businesses, abandoning password-only access. In addition, a number of states are expected to adopt some version of Europe’s strict General Data Protection Legislation. California, for one, has already passed legislation that will make it easier for consumers to sue companies after a data breach, starting in 2020.

The upshot is that individuals, businesses and government entities need to do everything possible to improve the state of their cybersecurity. They cannot eliminate breaches, but they can avert some and improve the chances of mitigating them.

IBM Africa and Hello Tractor pilot AI/blockchain agtech platform

IBM Research and agtech startup Hello Tractor have developed an AI and blockchain-driven platform for Africa’s farmers. The two companies will pilot the product in 2019 through an ongoing partnership co-financed by IBM.

Dubbed Digital Wallet in beta, the cloud-based service aims to support Hello Tractor’s business of connecting small-scale farmers to equipment and data analytics for better crop production.

“Agriculture is a complex industry that can have so many different variables. We’re bringing a decision tool to the Hello Tractor ecosystem powered by AI and blockchain,” Hello Tractor CEO Jehiel Oliver told TechCrunch.

The startup joined IBM Research to demo the new service at Startup Battlefield Africa in Lagos.

Available to Hello Tractor clients, the online platform will use a digital ledger and machine learning to capture, track, and share data, while “creating end-to-end trust and transparency across the agribusiness value chain,” according to an IBM release.

Digital Wallet will draw on remote and IoT-based weather-sensing methods and AI to help farmers determine crops and inputs, choose when to plant and optimize and predict crop yields.

The cloud-based dashboard also employs a blockchain ledger to improve multiple points of Hello Tractor’s business.

“We’re an agricultural technology company. Our platform connects farmers who need tractor services to tractor owners who own these assets as a business. We create that marketplace to bring supply and demand together,” said Oliver.

The demand stems from the 80 percent of Sub-Saharan Africa’s crops harvested without tractors or machinery and the 50 percent of the continent’s farmers who suffer post-harvest losses annually, according to IBM and the Food and Agricultural Organization.

IBM and Hello Tractor’s Digital Wallet will also loop in data from fleet owners regarding tractor use, track and predict repairs and servicing and build credit profiles to open bank financing for farmers.

Hello Tractor is a connecting service — neither the startup nor its farming clients own tractors. Founded in 2014, the venture began operations in Nigeria and has expanded into Kenya, Mozambique, Senegal, Tanzania and Bangladesh within the last year, according to its CEO. A for-profit entity, Hello Tractor has raised funding from private investors, DFI grants and a seed round.

The company currently generates revenue by selling the tractor-monitoring devices and software subscriptions for its app, according to Oliver. Hello Tractor doesn’t yet charge transactional fees for connecting tractors to farmers, “but we’ll be testing that next year,” he said.

The startup also plans to create broader revenue opportunities from data analytics.

“At this phase we focus primarily on mechanization, but coupling the insights being generated through that device with the IBM platform solutions specifically for agriculture can extend the value we offer our customers and…be monetized,” said Oliver.

He estimates the business of connecting small-scale farmers to tractors as a “multi-billion market” globally and pointed to Nigeria as the African nation with “the largest inventory of arable-uncultivated farmland,” 37 percent of the country, according to World Bank data.

IBM Research’s co-financing to build Digital Wallet does not include any equity stake in Hello Tractor, IBM confirmed.

The collaboration aligns with IBM’s global agricultural strategy, embedded largely in its Watson AI business platform and global agtech partnerships. As TechCrunch covered, IBM partnered with Kenyan agtech startup Twiga earlier this year to introduce to Twiga’s network of vendors a blockchain-enabled working capital platform.

IBM Research views the partnership “as scientific research collaboration,” according to VP Solomon Assefa.

“Through all its touch points — farmers, machinery, dealers, crop yields, data inputs — Hello Tractor is convening the whole agricultural ecosystem,” he said.

As discussed at Startup Battlefield Africa, Africa is shaping its own blockchain-focused startups and use cases — characterized more by utility than speculation. On the crypto-side, there were several 2018 ICOs, including remittance startup SurRemit’s $7 million token launch, payments venture Wala’s $1 million offering and one by South African solar energy startup Sun Exchange.

IBM Research and Hello Tractor teams will continue to build out the blockchain-enabled Digital Wallet on a lab, engineer and business level throughout 2019.

“We’re cultivating the partnership… including the executive and go-to-market side. You also have to focus on how you scale,” said Assefa.

Tesla’s China factory and the missed growth opportunity

Tesla made its ambition for world domination known when it announced its intention to build a factory in China. The move makes sense — China is the world’s largest automotive market. But it might be shortsighted.

By continuing to go after the higher tiers of an established market, Tesla will engage in a zero-sum game for market share instead of forging a new market of unparalleled size. Competition will be fierce as incumbents, like BMW and Audi, which are determined to hold on to their more profitable customers, respond in kind.

Instead, the larger opportunity for any automaker is to grow the overall market by way of disruptive innovation — and not in the Silicon-Valley-hype sense of the term. The architect of disruptive innovation, Harvard Business School Professor Clayton Christensen, explains that disruption happens at the low end of the market — not the end adorned with high-tech features and flashy designs.

Disruptive innovations succeed by transforming complicated and expensive products into simple and affordable ones, thereby enabling a much larger population to benefit from the offerings. And since they, by their very nature, expand the market, they constitute a wellspring of new growth.

Rather than take a page out of the disruptive playbook, Tesla is engaging in sustaining innovation. The company plans to use the new factory to build Model 3 and Model Y cars. Assuming that Tesla continues with its current positioning, these cars, like Tesla’s other models, will enter an established market to compete along existing measures of performance, like acceleration, style and luxury.

Sustaining innovations are important in that they advance an industry, but they offer little net growth, as not all consumers are able to access them. And because sustaining innovations target an industry’s more profitable consumers, we can expect leading automakers to fight tooth and nail to retain their core customers. Alternatively, a disruptive strategy offers a much easier way to tap into the Chinese market — and it’s already happening right under the noses of Tesla and other leading automakers.

Disruption happens at the low end of the market.

Chinese manufacturers of low-speed electric vehicles (LSEVs) — small vehicles that typically top out around 45 mph, have a limited driving range, and sell for as little as $2,000 — are creating a market where none existed, by primarily selling cars to people in rural China who have never owned one. We call these customers nonconsumers of cars. The measures of performance that matter most to nonconsumers aren’t speed, style or comfort, but rather affordability, accessibility and simplicity. So, as long as LSEVs meet these criteria, nonconsumers will generally be willing to buy them. After all, having a car that can’t travel very far or very fast is much better than the alternatives: bicycles, motorcycles or farm vehicles.

By targeting nonconsumers, LSEV manufacturers have steered clear of direct competition with incumbent automakers — who have at their disposal far more resources, such as capital, factories and relationships with suppliers — and effectively established a foothold that allows them to steadily move upmarket.

Taking the disruptive route has enabled LSEV manufacturers to unleash a new wave of growth that Tesla and other automakers should covet. During the decade that LSEVs have been available in China, sales have soared. According to the International Energy Agency’s “Global EV Outlook 2017” report, between 1.2 million and 1.5 million units were sold in China in 2016 — overshadowing the number of battery and plug-in hybrid electric cars sold globally that same year. Undoubtedly, further growth potential for LSEVs in China is immense — more than half a billion Chinese lived in rural areas in 2016.

Whether LSEV manufacturers manage to profitably march upmarket into higher-performance tiers of the market remains to be seen. What we can say for sure is that there is enormous untapped potential to be discovered — both in China and in other emerging markets.

Existential education error: Failing to train students on software

Although many of the milestones of the digital revolution have sprung directly from the research output of America’s colleges and universities, like Athena from Zeus’s forehead, on the instructional side, American higher education has taken a laid-back approach. Sure, there are more courses in computer science, millions of students taking courses online and MIT just committed $1 billion to build a new college for AI. But a campus-visiting time-traveler from 25 or 50 years ago would find a very familiar setting — with the possible exception of students more comfortable staring at their devices than maintaining eye contact.

This college stasis may be even more surprising to visitors from the transformed workplace. Jobs that made no or marginal use of digital devices 10 years ago now tether workers to their machines as closely as today’s students are glued to their smartphones. Processes that involved paper are now entirely digital. And experience with relevant function- and industry-specific business software is required in job descriptions for many entry-level jobs.

This hit home a few weeks ago when speaking to an audience of 250 college and university officials. I asked which of their schools provide any meaningful coursework in Salesforce, the No. 1 SaaS platform in American business.

Not one hand went up.

There are many reasons for this. Few if any faculty have dedicated their careers to (or even get marginally excited about) equipping students with the skills they need to secure and succeed in their first jobs. No one’s losing their job (yet) over failure to help students get jobs. Another is the cost of teaching; with strong employer demand for these skills, finding and hiring capable faculty costs more than teaching non-technical subjects. Finally, there’s the rapid pace of change in technology, and the sense that any educational effort will be obsolete in a few years. (Of course, the reality of business software is quite different; foundational platforms like Salesforce have a long shelf life — 10-plus years and counting — and some platforms are expected to last for a generation.)

But the primary reason colleges aren’t educating students on the software they need to launch their careers is the notion that it’s unnecessary because millennials (and now Gen Zers) are “digital natives.”

The idea of digital natives isn’t new. It’s been around for decades: Kids have grown up with digital technologies and so are adept at all things digital. It’s certainly true that today’s college students are proficient with Netflix and Spotify and smartphones. But it’s equally true that the smartphones they’ve grown up with haven’t remotely prepared them to use office phones, let alone career-critical business software.

Business software is really hard, even for digital natives.

Eleanor Cooper, co-founder of Pathstream, a startup partnering with higher education institutions to provide business software training, notes that millennials and Gen Zers are “accustomed to Instagram-like platforms which are both intuitive and instantly gratifying. But without exception, we find the user experience of learning business software to be exactly the opposite: instant friction and delayed gratification. Students first face an often multi-hour series of technical steps just to get the software set up before they begin working through tedious button-clicking instructions, which are at best mind-numbing and at worst outdated and inaccurate for the current version of the software.”

In an article in The New Yorker last month, “Why Doctors Hate Their Computers,” Dr. Atul Gawande describes the challenge of implementing Epic, a SaaS platform for managing patient care: “recording and communicating our medical observations, sending prescriptions to a patient’s pharmacy, ordering tests and scans, viewing results, scheduling surgery, sending insurance bills.”

First, there’s 16 hours of mandatory training. Gawande “did fine with the initial exercises, like looking up patients’ names and emergency contacts. When it came to viewing test results, though, things got complicated. There was a column of thirteen tabs on the left side of my screen, crowded with nearly identical terms: ‘chart review,’ ‘results review,’ ‘review flowsheet.’ We hadn’t even started learning how to enter information, and the fields revealed by each tab came with their own tools and nuances.”

Business software is really hard, even for digital natives. Today’s students are accustomed to simple interfaces. But simple interfaces are possible only when the function is simple, like messaging or selecting video entertainment. Today’s leading business software platforms don’t just manage a single function. They manage hundreds, if not thousands.

Gawande references a book by IBM engineer Frederick Brooks, The Mythical Man-Month, which sets forth a Darwinian theory of software evolution from a cool, easy-to-use program (“built by a few nerds for a few of their nerd friends” to perform a limited function), to a bigger program “product” that delivers more functionality to more people, to a “very uncool program system.” Gawande points to the example of Fluidity, a program written by a grad student to run simulations of small-scale fluid dynamics. Researchers loved it, and soon added code to perform new features. The software became more complex, harder to use and more restrictive.

And so beyond cumbersome interfaces, the second reason why business software is really hard is that it has become inextricably and tightly wound up with business processes. Salesforce consultants will tell you it’s easier to conform your business practices to Salesforce than to try to customize (or even configure) Salesforce to support the way you do business today. And that’s true for almost all business software. As Gawande notes, “as a program adapts and serves more people and more functions, it naturally requires tighter regulation. Software systems govern how we interact as groups, and that makes them unavoidably bureaucratic in nature.”

The myth of the digital native is convenient for colleges and universities, because it allows them to stay focused on what faculty want to teach rather than what students actually need to learn.

Software-defined business practices are increasingly standardized across functions and industries, and highly knowable. And because they’re knowable, hiring managers want to see candidates who know them. So it’s not just about educating students on software; inherent in preparing students on business software is equipping them with industry and/or job-function expertise. And that requires much more than 16 hours of training.

“Why can’t our work systems be like our smartphones — flexible, easy, customizable? The answer is that the two systems have different purposes,” Gawande explained. “Consumer technology is all about letting me be me. Technology for complex enterprises is about helping groups do what the members cannot easily do by themselves — work in coordination.”

The myth of the digital native is convenient for colleges and universities, because it allows them to stay focused on what faculty want to teach rather than what students actually need to learn. But it’s self-centered, superficial and silly. Rather than thinking about technology in terms of Netflix and smartphones, walk down the street and take a look at the software being utilized to manage your college’s admissions, financial aid and human resources functions. Indeed, 95 percent of your graduates will begin their careers working in places that look a lot more like this than like the faculty lounge. And that’s if they’re lucky. Otherwise they’ll begin their careers working in places that look a lot more like Starbucks.

In his article, Gawande notes that despite the many challenges of adapting to working (and living) on a business software platform, software is eating the world for a good reason: to improve outcomes for consumers. The Epic implementation should allow hospitals to scan records to identify patients who’ve been on opioids for more than three months in order to provide outreach and reduce risk of overdose, or to improve care for homeless patients by seeing that they’ve already had three negative TB tests and therefore don’t need to be isolated. “We think of this as a system for us and it’s not,” said the hospital system’s chief clinical officer. “It is for the patients.

These improved outcomes are synonymous with the data analytics revolution — a revolution that has colleges and universities excited about new programs and increased enrollment. But all the additional data to improve these outcomes needs to be captured first. And that’s done with complex business software. So it’s unfair, or at least hypocritical, of colleges and universities to attempt to pick the fruit of big data without first sowing the seeds. And sowing the seeds entails a serious investment in preparing students with the technical and business process knowledge they’ll need to use the software that makes big data possible.

Proof of use: A new crowdfunding threshold for passionate users

Right now, most participants in U.S. private placements must be “accredited” investors, meaning $200,000 annual income over multiple years or $1 million in net worth, not including your primary residence. These numbers have not changed since 1982, though inflation in the intervening decades has more than halved the real wealth they represent.

This means your mother, who owns a vacation home on Cape Cod, may be getting phone calls from boiler room broker-dealers. The wealth standard means your mother is considered qualified to evaluate such offers; a more sophisticated, but less wealthy individual is not.

Pending legislation addresses that. The Fair Investment Opportunities for Professional Experts Act is a revision to the 2014 JOBS Act. If you have a financial services license or are determined by the Securities and Exchange Commission (SEC) to have qualifying education or experience, it would allow you to invest, regardless of your wealth.

Proof of use versus proof of knowledge

The proof of knowledge approach is problematic. The advantage of the wealth-based standard for accreditation is it’s clear and straightforward.  A knowledge-based standard is more subjective, leading to potential disputes. Such a subjective standard may or may not open investment opportunities for people otherwise excluded by the wealth-based standard, but it’s sure to bring more revenue for lawyers.

Instead of a standard based on education or experience, the SEC ought to adopt a standard based on use — i.e. their contributions of time and talent that precede the investment. Call it “proof of use.”

The early growth of Facebook and internet protocols like HTML are analogies often used in crowdfunding and in crypto assets. If the volunteer developers who built the open-source internet protocols had been able to invest in them, they would be today’s internet moguls, alongside Jeff Bezos and Mark Zuckerberg. Meanwhile, early adopters, sellers and evangelists contributed tremendous value to Facebook and Amazon. If they had been able to invest, they’d have participated materially, too.

So-called “gig economy” platforms Airbnb and Uber have made similar recommendations to the SEC. Airbnb did so in a September letter to the SEC, advocating an update to rules governing equity-based employee compensation, which would allow them to distribute stock to hosts that use its platform without running up against the public-reporting limit of 1,999 shareholders. Uber had conversations in 2017 with the regulator, advocating similar changes for granting stock to its drivers. (In other venues, Uber has successfully argued its drivers are nothing like employees.)

I’m suggesting something a little different. If we are going to re-imagine a next generation of Facebooks that grow without information silos and monopolistic ambitions, network users must be able to contribute capital. Regulators could make this possible if they open the door for users who have been active with a product for a significant period of time to actually buy its stock.

Long-term, active use is a more objective standard than knowledge. Picking stocks based on individual experience as a professional or a consumer is also a time-honored principle, going back to one of Peter Lynch’s often-repeated mantras: “Invest in what you know.”

Civil’s failure

I’m not the first to suggest this. A journalism project called Civil used “proof of use” to describe how it would quiz would-be investors about journalism and their project, before allowing a purchase of the CVL token offered in an initial coin offering that sought to raise $24 million in a two-week crowdsale. The quiz and the other steps involved did not make it easy, even for veteran journalists.

Before selling any of their stock, Civil’s crowd-investors would have to take some action using the token. Voting on funding for a journalism project was one example offered (though this probably makes Civil’s token more like a security, not less). In this way, Civil hoped to be perhaps the first token crowdsale to legitimately demonstrate a so-called “utility token” exclusion from potentially applicable securities laws.

Civil’s crowdsale didn’t fail because of its self-imposed sophistication standard, or because the idea of a “utility token” is naïve in any business other than Chuck E. Cheese’s. It failed because it was trying to raise $24 million in two weeks for a community journalism project. It did raise $350,000, which to this former journalist sounds like a smashing success.

Real proof of use would be putting Civil’s $5 million seed round to work, demonstrating user traction — then opening its offering to its crowd of passionate users.

Proof of use = proof of users

Proof of use would have the added benefit of limiting the crowdfunding option only to companies that can actually attract users. Proof of that traction would be financed by wealthy investors who can bear the risk; for the growing company, proof of use would open a new financing option and a better path to reward its early adopters.

Right now, there is so much private money chasing deals, the best have no need to resort to crowdfunding. Broadening the accreditation standard only creates opportunities for bottom-tier deals or much less knowledgeable investors, or allows venture capitalists to front-run the entire crowd. This they already often do in crypto issuance, allowing their name on the deal to pump interest among retail investors who don’t realize they’re buying the opportunity at a higher price.

Proof of use would provide an additional fundraising avenue for products and services that are showing traction with users — one that would carry the added benefit of motivating the user base, besides the capital it brings. Right now, despite the billions raised by ICOs, users are scarce — only about 24,000 are active daily, across more than 2,000 decentralized applications, or Dapps.

I can think of three projects that have approached us at New Alchemy that would benefit from a reform like this. It would be a miracle if U.S. legislators and regulators were able to pull it through in time for their fundraises — which means they will likely exclude U.S. investors, again. I hope there will be better options for the next few that come along.