Sunday, February 26, 2017

Early Indications February 2017: B2B Websites a year later

A year ago this month I published the results of a survey of 100 business-to-business websites. (You can review those findings here.) I recently repeated the exercise, staying with the 100-company sample rather than expanding to 250 as I had planned. While there were no shocking surprises, year-over-year trend data suggests some unexpected 
macro directions.

I. What stayed the same

The companies that used video well (Corning, Haas machine tools, Bobcat) are still relatively lonely as front-runners. Given the complexity of many B2B applications, training and troubleshooting videos were relatively rare despite many potential use cases. (Schneider Electric has more than 200 vendor-neutral training videos and certifications in their “Energy University.”) While many companies post corporate overview videos, these were generally of little concrete help and long on a “Successories”-inspired aesthetic in both aural and visual look and feel.

Many good site architectures were the same — there’s no compelling reason why something that works well would need to be overhauled on an annual basis. Leaders here included Cisco, GE, Caterpillar, NXP semiconductors, and Rockwell Automation.

Very few websites included live chat, something I see as essential if a company is dealing with a) millennials or b) loud shop floor customer service scenarios. There was a modest rise in adoption over the year, but it’s still a tiny minority.

Customer value propositions were striking in their rarity. This bothered me more last year: in many cases this time around, the website is addressing so many constituencies (including investors, retirees, and recruits) that a customer focus may legitimately be more appropriate deeper in the site. I’ll say more about this later.

Many sites still presume a desktop user who can print long PDF product catalogs, rather than support mobile-first product data functionality. 

Most sites were organized by corporate org charts (industries, geographies, functional areas) rather than by customer questions. Similarly, many questions were answered by “call your sales representative or local dealer.” There are times when this process makes sense, but there are also instances in which the past default is carried forward solely by institutional inertia rather than using new technologies and behaviors to reinvent customer service from the outside looking in.

II. What changed?

Operational execution has improved. Many sites I had compiled into a list of “Don’ts”  were relaunched in 2016, while others got facelifts. Genuine howler-class errors were hard to find (minor chuckle: one paper company wished everyone a happy Chinese New Year in the year of the “roaster.”) Some sites haven’t been updated in nearly 15 years (hello Nutrasweet), but generally, content was relatively fresh, social media posts were rarely stale, and overall execution was solid, albeit from a Web-centric perspective. Only a small minority of sites (7%) were not mobile-friendly, down from a third last year, and of those 7, at least one (DHL) is in the midst of a site re-launch that addresses the mobile user quite well.

Apart from that, little changed dramatically in terms of site features, customer access, or integration with other marketing functions, particularly trade shows. The change that most interested me was more qualitative, a gradual migration of the online presence from being a storefront to being a broader window into the company. To oversimplify, I see a shift in many B2B sites from online commerce to digital business more broadly conceived. 

Part of this transition is a subtle shift in emphasis: plenty of companies have had “jobs” buttons or  tabs on their websites since day 1. Stock prices are not new additions to many front pages. Fedex has provided package tracking since the early days of their site. I do feel, however, that the best B2B websites I saw provided a richer representation of what the company _is_ (rather than “sells”) than in my past experience; your mileage may vary. One data point I’m watching: hardcore engineering-driven companies including GE and Parker Hannifin linked to Pinterest for the first time that I’ve seen.

This more holistic representation might be a function of the need to attract millennial recruits. In B2B especially, selling microprocessors, ball bearings, or industrial lubricants is not intrinsically appealing to people who typically lack familiarity with the industries and scenarios in which those products are used. Then to compound the difficulty of recruiting college students, many B2B websites were, to be charitable, limited in their digital adeptness. Thus students were being asked to help market products they didn’t use, understand, or relate to, using trade shows, newsletters, paper catalogs, and 50-something sales reps. Thus, my hypothesis goes, some of the increased use of social media, online video, and smartphone apps by B2B marketers is driven by a) younger staff and/or b) the recognition of the need to attract younger staff.

It’s no surprise that web presences appeal to investors: CEOs get paid not for selling widgets but for selling the stock. In chemicals in particular (Air Liquide, Dow, BASF, Henkel), I saw significant attention paid to investors, far more (at the top level of site organization) than to customers. In many more cases, the front page was a mixture of missions. Carried out well, such a heterogeneous approach gave a holistic sense of a company. Carried out poorly, it was design-by-committee without a central organizing design theme, organizational logic, or navigational rubric. 

An example of the former can be found at DB Schenker, a German logistics provider. News items related to the logistics industry, a corporate acquisition, and environmental impact all appear, alongside a conference invitation. The rotating photos “above the fold” relate to a new facility on the US-Mexico border, a smartphone app, careers, cargo insurance, and an outdated piece from November on holiday shipping and the retail sector. Thus multiple functions (public relations, investor relations, HR/recruiting, marketing, and sales) are represented, along with database connections to operations for shipment initiation and tracking.

In many ways, we are heading “back to the future,” to the e-business days of 20 years ago. Looking at B2B sites from 1996 in the Internet Archive’s Wayback Machine reminded me that much like today’s millennials, some companies were “born digital” while others migrated to this way of doing business. Thus Cisco’s embedding of business processes in digital backbones represents a different ethos, investment landscape, and demographic than does most B2B companies’ migration. As more paper/analog-based executives retire, as digital content platforms improve, as interactive agencies understand the uniqueness and complexity of B2B, and as mobility forces a re-architecting of many companies’ online presence, I believe the survey results in 2020 will look significantly different.

For all this conventional talk about a gradual evolution, however, there is major change afoot: many industries are embracing and/or struggling with the notion of embedded, networked sensors and actuators populating massive data stores of machine-to-machine traffic, the so-called “Internet of Things.” This new emphasis is probably the biggest change I saw between 2016 and 2017. It isn’t literally true that every company I surveyed had IoT on the front page, but it was WAY more common than just a year ago. Most accounting and consulting firms are highlighting it, as are industrial controls companies like Rockwell, infrastructure providers including Belden (which has an IoT phone conference scheduler on the front page) and power systems companies, hardware companies such as Cisco and Intel, and analytics software providers led by SAS and SAP. Thus market trends (along with the aforementioned demographics) will drive “digital nativism” deeper into the corporate hierarchy of many companies, demanding cross-functional collaboration, intelligent risk awareness and mitigation, new privacy policies and processes, and operational integration in something close to real time. Given the ever-expanding IoT market size estimates for 2020 (now in the hundreds of billions of dollars), I know what I’ll be watching for doing the next web census in 2018. 

Monday, January 30, 2017

Early Indications January 2017: Where’s the innovation?

As I was discussing the pace of change with my class recently, I struggled to name a hot young startup. It turns out there was a reason for that. Looking at the Fortune list of the biggest private companies with billion-dollar valuations, filtered for US head offices only, you get these companies, all with valuations over $5 billion:

Uber
Airbnb
Palantir
Snapchat
SpaceX
Pinterest
Dropbox
WeWork
Theranos
Intarcia Therapeutics
Lyft
Stripe.

Of these, Uber and Airbnb are of course interesting, and valuable, but it’s hard to call them tech startups, based as they are on the so-called sharing economy model. (Lyft is valued at about a ninth of Uber.) Palantir is an intelligence/defense contractor, so 99% of people won’t knowingly interact with it or recognize it. SpaceX is a literal moonshot, again, not really a typical tech startup. Pinterest feels like it could have been big, but given that its valuation is higher than a projected IPO, it feels like an underwater mortgage. WeWork is more like a REIT than an Apple or Google, Theranos is discredited and won’t likely be on the 2017 list, and the biotech firm Intarcia is almost fifteen years old, focused on a diabetes drug. Stripe builds payment infrastructure, a classic B2B play. That leaves one sole unicorn in the Netscape/Google/Facebook mold: Snapchat. An IPO there could draw some attention, but I can’t see it being a seismic event on par with Google or Facebook.

It also bears noting that none of these companies is at all young. Intarcia is 22 years old, while the youngest companies date from 2010-11. Put another way, here’s a list of important tech IPOs:

Apple 1980
Compaq 1983
Lotus 1983
EMC 1986
Microsoft 1986
Oracle 1986
Sun Microsystems 1986
Dell 1988
Electronic Arts 1989
Cisco 1990
AOL 1992
Netscape 1995
Yahoo 1996
Amazon 1997
Netflix 2002
Google 2004
Facebook 2012

Note the slowdown after 2000 in “blockbuster” IPOs of companies that return value over a relatively long span; such companies as Etsy, Fitbit, GoPro, Twitter and Zynga all have flopped after hitting the public markets. Tesla stock has performed well thus far despite never turning a profit, but that can’t last forever. Netflix required a lot of patience: if you bought at the IPO, it took 8 years for the stock to stop flat-lining, but since 2010, it’s risen from $8 a share to more than $140. All in all, we seem to be in a lull as far as fast-growing tech startups are concerned (with the caveat that Uber and Airbnb are both game-changers precisely because their asset model breaks traditional assumptions).

Several forces are at work, I’m hypothesizing:

1) The App Store platform model has lowered the barrier to entry for software developers. It’s hard to find a major pure-play software company of any magnitude in the past 10 years. The enterprise market has some counter-examples, to be sure: Workday, VMware, Palantir, and Tableau each have market niches, but none dominate an entire industry or have broad public visibility.

2) The gap (in the wrong direction) between private market valuations and public market outcomes is making many companies hesitate before launching an IPO. Dropbox has (or has had) a paper valuation of $10 billion. Its public traded competitor Box has a market cap of $2.2 billion on revenues of $300 million, so Dropbox would need to be pulling in roughly 4-5 times that — in the neighborhood of $1.5 billion — to justify such a lofty pre-IPO price. Staying private prevents that gap from becoming public, but it also delays the funding entities’ exits.

3) The next frontiers in computing — big data, AI, robotics, autonomous vehicles, Internets of Things — will often be capital-intensive ventures. It’s hard to see a startup outmaneuvering GE or Caterpillar on locomotive instrumentation, or disrupting Rolls Royce or Pratt & Whitney with revolutionary jet engine monitoring systems and software. Cloud plays like Rackspace, Cloudera, and Dropbox will be similarly asset-heavy, making Facebook or Google-like multiples difficult now that the industry is both mature (in its tight margins) and operating at huge scale. Meanwhile, most IoT or autonomous vehicle operations will need deep pockets: Uber bought Otto, Google’s car business (Waymo) finally has a name to go with its budget, and the incumbents (Volkswagen/Audi, Toyota, GM, Ford, Delphi, Continental, Bosch) are busy as well. Factors like product liability can quickly discourage garage-scale operations, as they did George Hotz’s Comma.ai effort last October. Robotics isn’t a big factor in the unicorn list; maybe that will come later.

4) Starting with Netscape, web-based software businesses have had a difficult time getting money from retail customers, compared with Lotus, for example. Netflix is the rare content play that turns a profit from direct payments; AOL, news media, standalone music services, and even investment advice sites are struggling. Some have tried the enterprise route, but the big successes have been ad-funded. Given the enormous power (speaking here of the U.S. market) of Google and Facebook, it’s hard to see how Snapchat, Vine, Twitter, Tumblr, or some new startup can break into that select club. Even Quora, with its vast knowledge base, seems content to run low-key ads that likely don’t pay the rent. On the demand side meanwhile, people are accustomed to getting good stuff for free (“consumer surplus,” in economists’ terms), making the ad model viable and in many sectors essential. There are only so many hours of human attention in a day, though, and getting new share means dislodging some well-entrenched incumbents.

5) Maybe, in line with what the economist Tyler Cowen has argued, a broader innovation slowdown is hitting the tech sector. When you look at our grandparents or great-grandparents, some of whom lived through both the Wright Brothers’ first flight and the 1969 moon landing, mass electrification and the atom bomb, penicillin’s introduction and the MRI, open-heart surgery and test-tube babies, the Internet’s origination and the first cell phone, 1900-1980 was a period of innovations that reshaped everyday life. Since 1980, what else is in that league besides the smartphone and World Wide Web, obviously? Fracking transformed oil and gas, the mini-mill reinvented the steel industry, and minimally invasive surgery is the norm for many procedures (as are stents rather than open-heart surgery). Cloud computing is reshaping the server and now storage markets, Skype was revolutionary (but non-revenue-producing) before Microsoft tamed it, and Google search solved a very hard technical problem. GPS reinvents our sense of space and location. But will we really look back on Facebook, YouTube, and LinkedIn on the same plane as the automobile, television, or the transistor?

6) Speaking of the smartphone, the final factor affecting our perception of innovation is the globalization of tech. Whether it’s the Japanese messaging app Line raising $1 billion last summer, Alibaba’s record-setting $25 billion offering in 2014, or privately held Xiaomi’s status as a pre-IPO hardware company worth $45 billion, the biggest stories are all global plays (Uber and Airbnb among them). More and more are headquartered closer to the fastest growing markets and/or talent bases outside Silicon Valley: India’s Flipkart (an online retailer), Sweden’s Spotify, Coupon (a South Korea e-commerce business), or Global Fashion Group (an e-business focused on apparel serving 24 countries based in London). Innovators can be literally anywhere, building apps and businesses North Americans never see or even hear of.

So is innovation slowing own? I think it makes sense to set a baseline: a huge percentage of human codified knowledge is now online, often for free. Many, soon most, adults on the planet have a networked supercomputer close at hand, often in a pocket or purse. Everyone with these devices knows exactly where he or she is at any time; can reach millions or billions with a tweet, a post, or a blog; and can capture and watch high-definition video and still images. That’s our baseline of “interesting,” which has to count for something. At the same time, we still burn coal and petroleum for most of our mobility and much of our illumination, train service is pathetic in much of the world, and human life expectancy extension may be slowing down or even reversing. In the narrow realm of content, E-book sales are slowing, vinyl LP sales are expanding rapidly off a very small base, and even cassette tapes are in favor among some hip populations. 

Do we measure innovation by the magnitude of problems we have solved, or by the frontiers left relatively unconquered? Is the IPO success of an online merchant important for quality of life, relative to the possibilities for telemedicine or Kenya’s mobile banking success? As usual, “it depends” sounds like a copout, and maybe it is, but I do long for the days when hardware, software, and services for the “average” North American were new, exciting, and a bit rough around the edges as compared to the tech landscape of the media and entertainment period we currently inhabit.