Origami Around
Three Goblin Art
❣ Chile in a Photography ❣
d e v o n
No title available
🪼
JVL
Product Placement
@theartofmadeline
Stranger Things
h
"I'm Dorothy Gale from Kansas"
Love Begins
No title available
Aqua Utopia｜海の底で記憶を紡ぐ
ellievsbear
Alisa U Zemlji Chuda
noise dept.
I'd rather be in outer space 🛸
#extradirty
seen from South Korea
seen from Lithuania
seen from Brazil
seen from United States
seen from Brazil
seen from United Arab Emirates
seen from Argentina
seen from Sweden
seen from United States
seen from United States
seen from United States
seen from United States
seen from United States
seen from United States
seen from United States
seen from Malaysia
seen from United States
seen from United States
seen from Brazil
seen from United States
@nicpoulos-blog
5 Opportunities In Sales Enablement
Sales enablement tools and the startups building them have taken on a new level of importance over the last few years as sales teams have become increasingly quantified and metrics-driven. While larger players like SalesForce (SalesCloud) and InsideSales.com have build successful sales enablement businesses taking a platform or suite-oriented approach, most emerging solutions have been point…
View On WordPress
4 Reasons Enterprise Analytics Will Define 2015
In 2015, Enterprise Analytics will not only attract more corporate spend, but it is also poised to be a major driver of high-growth technology M&A going forward. Enterprise Analytics solutions are finally proving out their long-touted promise of eliminating waste and improving business performance in large organizations by crunching data and delivering actionable to-do’s in a way (or at a speed)…
View On WordPress
Vertical SaaS: What Is It & How Is It Different?
Vertical SaaS: What Is It & How Is It Different?
Vertical SaaS has become a hot topic of conversation in the tech, startup and venture capital world over the two years. Big IPO stories like Guidewire, Opentable, Fleetmatics, Textura and (probably most notably) Veeva Systems have helped fuel the fire, as have the increasingly large pools of capital attracted by startups like Kenandy, Procore, Mindbody, CareCloud, and Doximity. Here at Bowery,…
View On WordPress
CRM Data All Sales Teams Should Track
CRM Data All Sales Teams Should Track
Last week, Joe Caprio of InsightSquared joined us in the studio to record our 21st episode of the Bowery Capital Startup Sales Podcast: “Must-Have CRM Fields For Data-Driven Sales Teams.” In this session, we walked through the top CRM data fields that Joe feels have really allowed his sales org to streamline its funnel, highlighting a few that are less commonplace. In this follow-up post, I’ll…
View On WordPress
Enterprise Innovation Through Startup Partnerships
Enterprise Innovation Through Startup Partnerships
Enterprise innovation has been a topic of academic focus since well before the MBA existed but only in recent decades has it become such an imperative. In 1997, Clayton Christensen introduced the world to his concept of the “Innovator’s Dilemma,” wherein large companies that focused too much on customer needs of today (versus their future needs) would miss out on game-changing products and…
View On WordPress
Big Data Wildcatting
As posted earlier on Square1's blog and the BoweryCap blog.
From its debut in the late 1980’s, the term “Big Data”—like its lexical cousins Big Oil, Big Tobacco and Big Pharma—was engineered to intimidate. Steve Lohr of the New York Times wrote a wonderful account of its origin as a phrase just last year. One originator was author Erik Larson —of Devil in the White City fame—who lamented the adverts piling in his mailbox (the original spam some might say). “The keepers of big data,” Larson wrote in a 1989 Harper’s Magazine article, “say they do it for the consumer’s benefit. But data have a way of being used for purposes other than originally intended.” Another originator—John Mashey of early SFX giant Silicon Graphics—used the term to educate potential customers on the wave of digital information to come. “[What matters is] what you do with it,” Mashey wrote in one presentation , “but if you don’t understand the trends it’s what it will do to you.” In both cases, and probably most others early on, Big Data was a colloquialism meant to convey the magnitude of measurable info that companies would inevitably have to deal with in coming years.
In the last two decades, the sense of urgency around Big Data, especially as directed toward business decision-makers, has been amplified: recent cover stories warn of the “ Data Deluge ,” announce that “ Data Is Power ,” or—in a most dramatic case—ask “ Is Data The New God?” Most of these treatments do a great job of impressing the importance of taking action around Big Data but give little insight as to how to do so. Only recently have I seen smart publications raise an equally worthwhile consideration: “ You May Not Need Big Data After All .” The most important audiences in the room—the execs, entrepreneurs and business users—need an impartial framework for thinking about leveraging the digital information their organization does or may have access to. By now, execs should understand the dangers of inaction around emerging tech trends, from loss of market share to competitors to leaving cash on the table. But we should keep in mind that committing prematurely to open-ended and / or cash-intensive “Big Data strategies” can be equally as crippling.
Rather than view Big Data as an obstacle to be tackled or treasure to be won, I suggest we think about it as a natural resource, in the same way we would timber, wind or fossil fuel precursors. This framework helps us distinguish between two very different conversations: (a) what raw signals are accessible that may be useful? and (b) what are the costs and benefits to extracting it into useable information?
The volume of measurable raw signals is exploding exponentially thanks to unprecedented levels of web traffic, exploding mobile usage both at home and work, and the proliferation of a wide range of connected devices. To give you an accessible reference point, the identifiable storage capacity of the human brain is around 2 Petabytes . That's about 4,000 Macbook Airs worth of capacity. Facebook , Microsoft , Google and Amazon , amongst others, all boast data warehouses equivalent to hundreds of brains. The amount of digital information generated in the last two years represents the overwhelming majority—likely to the tune of 95%+—of extant data today.
All of those parties, however, devote massive resources to “mining” the raw signals that customers generate in their wake, “refining” them into a workable format and storing the results in an orderly fashion. Depending on the business, this process of extracting useable info from raw signals can be complex and expensive, just as is turning sedimentary petroleum into car-ready gasoline. For example, telecom companies all have some level of access—whether they parse it or not—to customers’ behaviors pre-purchase, while active subscribers and post-cancellation. Solutions exist that can turn those behaviors into insights that will help the telecoms win more customers or prevent losing them. Many software startups have attracted millions in investment dollars to provide such customer analytics; core competencies amongst them vary, but Gainsight , BluenoseAnalytics , Totango and ActionIQ are a few relevant examples. As customers of these and other solutions have found, the potential benefits of a successful business analytics initiative are huge. Many other software vendors are working hard to solve companies' Big Data issues for them regardless of industry: Real Estate ( SmartZip ), Life Sciences ( Zephyr Health ), Government ( Premise ), Human Resources ( HiQ ), Fashion ( StyleSage ), Agriculture ( 640Labs ), and Transportation ( Transfix ), to name just a handful. Given the potential expense and management focus required in all cases, however, today’s executives must systematically determine which Big Data resources they can extract in an ROI-positive manner. 1) What raw signals does your business have ready access to? Like landholders that discover oil on their fields, search engines found early on that they could readily collect a wealth of signals on their customers’ behavior (thanks to their volume and high level of product interaction). Ad targeting was just one revenue-generating result. In some cases, it may also make sense to consider acquiring additional data access from 3rd parties to supplement core in-house data resources.
2) How complex will the process be to refine those raw signals into useable information? Identify your low-hanging fruit. Point-of-sales purchases are more straightforward than real-time video streams to collect, for example, because the latter usually requires more advanced database technologies that enable the querying of massive unstructured datasets. Moreover, the mere act of capturing data will result in some sort of expense; they are of little value without the ability also to reap actionable business insights from them.
3) What investments will you need to make to execute on this refinement process? A combination of data scientists and front-end business intelligence software will likely be needed to take full advantage of your natural data resource. Unlike most mining or energy companies, many Big Data adopters won’t need as much in the way of CapEx (thanks to As-A-Service providers and human capital). The investment sizes, however, can be just as significant and likely require different business planning (e.g. they may not be amortizable).
4) Once the refined insights are generated, what’s the true end value to the business? Oil and timber are easy to sell on the open market, but Big Data insights—in most cases—are of greatest value to the business internally (and privacy or competitive concerns may present regardless). Like any natural resource, Big Data may generate 10-figure revenue opportunities for some, while others will find it makes sense to invest on a limited basis or outsource more of the value chain.
In the modern enterprise, Big Data is a ubiquitous natural resource, completely renewable but only of potential value on the outset. Forward-thinking leaders must now commit to understanding the full complexity of their unique extraction process. ROI will vary by industry and business model. But a dedication to testing new, innovative strategies on a continuous basis—Big Data wildcatting , in a sense—is a core competency that all next-generation executives must now embrace.
3 Big Issues In HCM: 2014
Below, compiled from some recent reactions to industry developments, and colored with some excellent ideas from founders I’ve met in the space, are three issues in enterprise HR that I believe can and will be solved through HCM software in the near term. If you have any reactions to the below as you experience them in your organization, are founding a company to solve any of these problems, or know of any great tools that address them, I’d love to hear from you:
1. Career advancement / restructuring — Comparison of performance and role / salary / comp structure to make sure that employees aren’t being undervalued. Imagine something that would automatically provide a list of suggested promotions / demotions based on corporate priorities vs. actuals. We’ve been doing this for years on a high level with financials (e.g. where are the shortfalls and why). Moving closer to the source, optimization and A/B testing are now helping us to make good on these opportunities. And we are already starting to do this in the planning / internal phase of things, namely via new DevOps practices and finally directly within Product Management groups (cf. startups like Wizeline, which is full disclosure one of ours). I posit large companies have huge invisible costs tied up in the gap between comp structures (based on qualifications / original expectations of performance ability upon hiring) and actual performance shortfalls due to structural inefficiencies.
2. Identify employees at risk of plateau / stagnation / flight — Turnover is a huge invisible cost for organizations, one I’ve written about before. If you could identify specific cases and likelihood accurately, this solution alone would be a potentially big software product. E.g. Joe is a 85% likely flight risk in the next 6 months due to various factors (e.g. performance, last promotion, salary vs. peers, commute time, marriage status, etc.). A nod to Darren Kaplan of HiQ, a great startup addressing this very issue. To some extent this is a data collection issue and solving it properly likely requires ML expertise (def. data-science heavy at the least). Ever work for a huge company, or know of someone who companied of “lack of visibility” or rigid management structures? If so I bet you know this problem first-hand.
3. Improve collaboration + influence intra-organization — This is a bit more of a vague one; tougher to determine what the “hard ROI” would be. But think about the proliferation of enterprise communications data we only recently have access to that’s going relatively unused. Theoretical example: John and Stacy are structured in the org together but only communicate 50% of the average. Or, looking at X company, outperformers email 100 times a day, while underperformers email 500 times a day (implication: initiative to reduce email). Also figuring out a way to make great / more experienced employees a resource for others (e.g. training young employees, solving point problems, for idea generation, or even just to maintain a healthy culture).
Don’t Always Be Closing: BDRs & Sales Role Specialization
BDRs, or Business Development Reps, are critical members of most modern B2B sales teams. Specific titles, processes and funnels vary by company, but in general it’s the BDR’s job to work with leads to get meetings. They start with the very top of the funnel: Marketing Captured / Qualified Leads (MCL / MQLs), those who come into the top of the funnel into some sort of leads list or CRM (e.g. through a piece of content marketing, free trial, webinar, inbound emails, cold outreach, paid lists, 3rd party lead gen services, etc.). And they need to turn these leads into Sales Qualified Leads (SQLs) that sales reps can then take and close.
The BDR role is relatively new but has become invaluable for three major reasons. (1) We’ve come to recognize that–like in so many businesses since Henry Ford’s assembly belts cranked out 93-minute Model Ts–specialization is the key to efficiency. The addition of new intermediary roles between lead and close recognizes that buyers of enterprise software move through various measurable stages of learning and vetting to get to true intent, and that we must treat them differently. During our last CRO Summit, David Skok of Matrix Partners articulated this brilliantly via an example I’ll paraphrase: “When you stroll into a general clothing store to browse and a sales rep won’t leave you alone, that’s incredibly annoying. It would be equally annoying if you walked into a tuxedo shop with a pressing need for wedding attire and couldn’t find one sales rep to help you.” (2) If you don’t define specific, distinct roles for your sales people, they will gravitate toward what they like, which may not align with what is most productive (another point David highlighted). Very few people prefer outbound prospecting and cold-calling, and provided a choice between that and inbound, most will choose the latter. If you hire, train and compensate people specifically to cold call, however, they won’t be surprised by the assignment will likely be better suited to it. (3) Finally, and most obviously, the more time anyone spends doing one thing, the better they become at it (especially when their comp is on the line). And often the better they are at something, they more they enjoy it. I couldn’t tell you how long it would take a BDR to master their craft, but I can say they can go from zero experience to a damn good performer in a fraction of Malcolm Gladwell’s 10k hours (more on his spurious claims here if you’re interested).
But what actually makes a good BDR? Again, think about the nature of specialization. BDRs are Sales Reps. So why would the same approach of selling the product as if you were about to close a deal work well for BDRs? The general consensus from several of the speakers who joined us at our CRO Summit 2 weeks ago was loud and clear: selling your product off the bat can be the worst mistake you can make as a BDR (and maybe even as a Sales Rep who’s just received an SQL). A quick anecdote Mr. Skok shared: Aaron Ross once sent 200 emails to F500 C-level execs: 100 “salesy” emails and 100 emails simply asking “who’s the right person in your company for X” (in that case, “X” was CRM). He got 10 responses: all from the latter “simple ask” group. NOT ONE of the “salesy” emails converted because the direct and simple ask, without the feeling of a sales obligation, was easy to respond to, whereas moving directly into a sales relationship both comes off as spam and puts people on guard. When he reached out to the contact referred within the org, he’d already established a level of trust given the intro was warmer (and likely from a superior), resulting in more meetings, higher conversions, etc.
Even when it comes to emails, as our panelists pointed out, good BDRs should assume that the target (a) does not trust you and (b) may not even have much purchasing intent when you reach them. To avoid being thought of as human spam, while hopefully helping your target move into / through the learning phase of the process, focus on the things that will add value to the recipients’ day. They probably don’t know your product, and may not even be aware of the problem your product is solving. But they do know when someone is unilaterally helping them. Once they feel an element trust and that the goodwill has been one-sided, learning more about a product that’s relevant and that could actually benefit them isn’t so much to ask (and it should assuming some decent pre-qualification, at least to the MQL level).
Instead of writing an intro email with a thick paragraph on your product as a BDR, find anything about that person that they will be interested in to lead before getting to the brief and direct ask:
- Content that will be of high value to the individual
- Information about a customer / contact you have within their Company
- Company- / Industry-specific news
- Other customers your Company has within their industry
- Great outcome you enabled for a competitor
- Reaction to a great blog post / tweet from the individual
- Anything personal about them at all that’s public / not creepy
- Share a piece of news that is relevant to their business
- Offer to do anything at all, even just follow up later, then do it (something as simple as this builds trust!)
Great sales teams understand that 7-9 email touches tend to be the magic number for optimal conversion. So you don’t need to be selling product from square one to be doing it effectively near to close. That said, for BDRs, it always pays to be direct. Don’t let “rotten” leads make it past you and make sure your company keeps an eye on its scoring methodologies so you have a programmatic way of sniffing these out. Outbound emails should always be simple and include the ask as clearly as possible; ideally within the first two sentences. A good model might be a awareness opener (“we just ran an amazing campaign with [competitor]”) then a clear ask for a meeting (“how’s Thursday at 9 AM?”). An ignored, rejected or delayed request for a meeting doesn’t always end a conversation, but neglecting to ask will fail to progress that portion of leads who would’ve accepted the appointment 100% of the time.
Glengarry Glen Ross coined the ABC sales mantra of “always be closing.” That general mentality of aggression is timeless. But in the new world of B2B sales, division of labor, specialization and quantification are the keys; one-size-fits-all cultures–be it regarding sales hiring, training, development, comp–are a thing of the past.
PS: Earlier this summer, we kicked off Bowery Capital’s 2nd annual CRO Summit with two segments that inspired the above post. We begin with a panel on “Improving Lead Gen With Unconventional Methods” for startup salespeople. Moderated by SalesHacker Max Altschuler, panelists included Javier Rosas (Head of Sales, Mag+), James Nielsen (Head of Sales, Wizeline), and Jake Dunlap (Founder & CEO, Skaled). That segment was followed up by a great tactical chat from David Skok, General Partner at Matrix Partners (VC) on Business Development Reps (BDRs) and lead flow. Thanks to all speakers, panelists, moderators and other participants for making this event our 6th quarterly CXO event the best yet!
50+ Tools For Startup Job Seekers (Public List)
From the Bowery Blog, posted earlier this morning.
One of the more common questions we get from friends in the tech community (including portfolio companies looking to recruit): what are the best resources for startup jobs and candidates? Over time, we’ve run across the full array of job boards, matching services, and more recently, apps that help people find startup roles. I started collecting a list a few months ago. Today, I’d like to share that list. It’s an open Google Sheet and I hope you’ll add any other services or tools you find worthy:
The List — Top Resources For Startup Job Seekers
For those of you in the job hunt, you’ll be pleased to know that more efficient and intelligent tools for sourcing candidates (or finding openings) are on the rise at a steady clip these days. Here are a few thoughts on why. First off, enterprises spend a lot on Talent Management systems and sourcing (e.g. ATS + outward-facing tools like LinkedIn Recruiter, Hired, etc. + headhunting services). Depending on the segment you’re looking at, most sources (like this one from Bersin by Deloitte) list a number around $5B. Another factor is that in comparison to other areas of software procurement, HR is a field that necessitates somewhat of an early-adopter attitude; naturally startups and tech companies lead the pack but even F500′s are used to packing on the HR tech subscriptions (Bersin estimates that the average number of HCM products in use at a large company is seven!).
In addition, enterprises are warming up to the fact that they have more Human Capital-related data than they know what to do with, and that they should make use of it. Instead of loading a static resume into Brassring (IBM Kenexa), companies can amass credentials data from both employees and candidates (the latter thanks in large part to LinkedIn). The next step is using that data to make hiring (and also performance / compensation management) smarter. Of late, large HCM players are facilitating this trend by focusing significant resources on middleware to allow for across-the-stack integration with other HR cloud platforms (as SaaS models have nearly taken over in this market), as well as proprietary analytics packages (e.g. Workday BigData Analytics, SuccesFactors Workforce Analytics, Oracle OBIE, SumTotal elixHR, Peoplefluent Collosus, etc.; more on those here, again, from Bersin).
Finally–and probably most relevant to you, job seeker–consumer UX best practices are finally coming to recruiting. More mobile-first interfaces, fewer endless online forms, less uploading of PDF resumes, more quality opportunities surfaced via push vs. pull. More than ever, hiring experiences are tailored to the candidate, with vertically / functionally focused on the rise (e.g. Proven for restaurants, HirePurpose for vets, Elegant.ly for designers, FirstJob for recent grads, etc.). That said, the hiring process isn’t quite set & forget yet; until then, keep an eye on the latest options and don’t hesitate to add any we missed. Good luck!
The Ultimate Guide To Startup Sales Tools
Repost from the Bowery Blog last week.
Bowery Capital - Guide To Startup Sales Tools presentation is live on SlideShare.
Here at Bowery Capital, we accelerate our portfolio companies’ growth by focusing squarely on support around early customer acquisition. We’ve helped many B2B startups build & scale sales teams and know that sometimes you need the right tools to win. So we thought it would be helpful to aggregate some of our team’s and founders’ learnings around early-stage sales tools that actually work. We hope startups will find this resulting guide to be helpful and maybe also learn a thing or two about how Bowery Capital thinks about B2B startups.
In the coming weeks, I’ll be releasing a set of blog posts walking you through how we view the landscape of sales tools (from pre- to post-sales) and what we think a founders’ key considerations should be in evaluating each. I’d also point out that the Bowery team is not only on the demand side of the equation (e.g. using these tools directly ourselves or seeing our startups do so), but also the “supply” side, in the sense that we invest in and help grow B2B business software businesses. So I’ll also provide a view on trends in the sales software market and where we feel it’s headed as the next generation of CROs looks to build a more powerful toolkit.
Finally, I’d like to emphasize that while we’ve worked hard on this first cut of the guide, we’d love for it to be a living document that can grow / improve over time. To that end, please reach out via Twitter, LinkedIn or email with any feedback. Think we messed up a price point? Hate a pro or a con? Feel we’re missing your favorite tool or startup? Please let us know so we can fold your thoughts into the next update!
The Unraveling Of Moore's Law
Since about 1980 we’ve gotten used to a ridiculous pace of technological innovation. At least as concerns processing power, it’s likely we just can’t keep up this pace of advance without a near-reinvention of the modern CPU. Chew on this: between then and 2005, the speed of available PC processors increased roughly 1000x.
Just compare that to other “indicators of progress” in our world: we definitely didn’t get one thousand times healthier (~0.29x fewer US deaths / 100k population ’80-’85 per the CDC), wealthier (~1.75x greater mean US net worth ’04 vs. ’83 per the Atlantic) or more numerous (world population ’80-‘05 via Wolfram Alpha: ~1.46x) over that period. Personal computers and the smart consumer electronics that followed have all benefitted from this Golden Era of processing power expansion. Moore’s Law—the predicted annual doubling of transistor counts—has become an expectation.
Unfortunately, Moore’s Law, which has already run into some bumps in the road, is widely expected to break down. But the good fight continues. DARPA, for example, recently launched an initiative to build a sub-20MW exascale computer, (1000 petaflops vs. our current max of about 50. Semis are by no means my area of expertise but I’m guessing there will be some interesting opportunities for innovation here. In any case, DARPA has done a few cool things in the past, like creating the internet, but my money would be on either a startup or a large consortium of tech companies (polar opposites, but hey, this stuff takes really big and patient money). A point for context: Q1 2014 saw 36 deals in Chips / Semis, the most in 2 years (representing $270MM in capital invested). For the sake of the innovation hungry consumer, good luck to them all!
PS: As I mentioned, this isn’t my area of expertise. But I’d love to hear anyone’s thoughts on which areas of innovation might be interesting here. Give me a shout: @picnoulos.
Back To The Future: The Rise Of Push Analytics
The term “predictive analytics” is grossly overused in the startup world these days. I’d say about 20% of the pitch decks I see reference some sort of predictive or machine-learning feature. Often it’s used in an aspirational way (e.g. on the product roadmap 3 years out) or really just kind of thrown in to add a data-driven aspect to an otherwise traditional approach. Investors, the press and industry types are in on it too. Using Google Trends data as a proxy, public interest in predictive analytics has been on a steady uptick, increasing fivefold since 2009.
We’re warming to the term “predictive” because it’s a handy catchall for everything taking advantage of a pervasive trend: leveraging the availability of tons of new data, many of which are real-time and some of which aren’t even structured (i.e analysis-ready). “Predictive” is a common thread between Amazon / Netflix recommendations, Palantir’s signal-from-the noise tech, Rocketfuel’s ad optimization magic, IBM Watson’s Ken Jennings-beating secret sauce, Siri (and Cortana!), and so on. Even President Obama got in on the excitement with a recent announcement of public & private funding for tools to support neural network mapping, equally neuro- and computer-science. The accurate prediction of complex outcomes—by systems so sophisticated as to appear “intelligent”—has been a holy grail of futurists for centuries. The contemporary prospect of harnessing “big data” in combination with new machine learning schema would seem to be a turning point that we’re only now approaching. The truth is, we still face a lot of hard work to make sense the growing pile of data we’re now able to collect about ourselves, our businesses and the world around us. But the progress we’ve made puts us on a pretty exciting trajectory:
Data Analysis For Prediction Is Ancient, But The Input Required Was Lame
Humans have been able to observe patterns within complex systems in order to accurately presage events for millennia. Several predictions we’ve become quite good at involve natural events like sunrises, tides and moon phases. Several ancient astronomers did do through dogged observation, literally marking the movement of astral bodies daily for decades in order to understand seemingly mysterious events like the retrograde motion of Mars or the transit of Venus. When the first automatic computers came along, the were thought of as time savers, or tools that merely sped up what man could already conceptualize (and even do himself given the time). Examples include the abacus, the proposed Babbage engine and punch-card mainframes. We did the real thinking, machines did the grunt work, then we pulled insights from their output.
Machines! They Learned Quickly, But Still Needed Tons Of Manual Feeding
In the late 1950’s, a new concept began to emerge around what computer “programs” (which began as little more than overgrows punch cards) might actually be able to do: make “intelligent” decisions rather than merely execute calculations. The term machine learning was coined in 1959. The basic gist was that computers’ ability to do X could improve autonomously, without human intervention, by “learning.” We’ll illustrate with a theoretical program whose task it is to distinguish a human face from other animals’ faces. To begin with, a programmer would “teach” the program a thing or two about the general characteristics of the human face vs. others animal faces. He’d do this by giving it a data set of many faces (of both types) plus the correct answers (human or not human). Then, the program would analyze the data and identify the graphical similarities of the human group, the similarities of the non-human group, and the most common differentiating factors, for example. Now, having learned from its “training set,” the program can be served new, unseen photos, and make independent determinations on whether or not it depicts a human visage. Over time, whether through additional training or continuous feedback, the predictive software should become better and better at its job. But at the end of the day, the decisions the program could make were still pretty one-dimensional, or limited by having to hand-choose factors & feed training data for each analysis.
Ubiquitous Data + Autonomous Learning = Hands-Free Intelligence
Since the inception of machine learning as a concept over 50 years ago, the ability of software to crunch more data faster has skyrocketed, thanks both to the work of pioneering data scientists and the steady forward-march of processor power. It’s gotten to the point where machines can run calculations so complicated that that they begin to appear “intelligent.” John McCarthy (a father of machine learning alongside Minsky, Shannon, Solomonoff, etc.) said it best: “every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.” If and when these simulations constitute human-grade “intelligence” is the real sticking point (and whether or not tests like Turing’s famous thought experiment constitute the best measures of success). Miles Brundageand Joanna Bryson wrote a nice piece on the topic, published not too long ago on Slate. They pose that we’d be remiss to downplay progress of technologies like Watson as merely computation, while highlighting views that we’re a far cry from AI (the stance of one of my favorite modern thinkers, Douglas Hofstadter). I’d tend to agree while pointing out that as the amount of data accessible grows, and our programs’ ability to learn from that data in an ad hoc manner expands, software begins to surface more “proactive analytics,” which may begin to blur the line between what we call predictive and intelligent.
What To Expect In Software: The Transition From Pull To Push
I think we’re likely now entering a 10-year period in which predictive analytics turn the corner from being a “pull” technology (e.g. dump in the data, get a result) and develop the capability to parse a large amount & variety of data to “push” the right insights, at the right time, in the right way to users. While it’s not necessarily intelligence, as these systems are able to access a variety & volume of data not easily comprehensible to humans, and analyze them with methods far beyond our computational abilities, they’ll appear to know something we don’t. This concept is certainly applicable to both consumer and business software, but in B2B, we’re already starting to see some pretty amazing applications.
Emerging software startups are attempting to, for example, identify the next likely location of a crime (Knightscope) or the best combo of lifestyle & meds (Lumiata) on a person-by-person basis. No one would give up their cops or doctors for the time being but one can begin to see how the line between what we deem predictive and intelligent is beginning to blur. In enterprise software, the reigns have been fully handed over to machines as far as some decisions are concerned. The proliferation of sensor data is revolutionizing logistics & SCM, increasing transit efficiency & inventory availability by 30-40% generally (and much more in some cases). Prices for tickets, hotels and other items (non-retail) online are often fully controlled by algorithms, which take advantage of weather, device, macro-economic metrics, etc.; as you likely know from Kayak, their results are inscrutable to the layman. Marketing is experiencing a complete decade-long transformation from a gut and style business (think Mad Men) to a fully quantified field advanced by data scientists and hackers (RocketFuel on the backend, and emerging players like TrackMaven in the space of marketer-facing applications) [disclosure: TrackMaven is one of ours!]. While it’s my view that we’re seeing most of this predictive analytics progress in the enterprise, expect it to enter your home and daily life if it hasn’t already. The key consideration is that in all these cases, complex & actionable insights are being surfaced without a commensurate increase in the work required to feed programs training data. Not to downplay the difficult work of software engineers here, but when our programs determine not only results, but when and where we need them, analytics becomes a much more frictionless “push” process that we can spend less dedicated time governing. And for the scientist, salesperson, marketer and consumer alike, that’s a big deal.
Now, the moment you’ve been waiting for: “where we’re going, we don’t need roads.” Or at the very least we won’t need to look up directions.
Stress In Startups: 5 Pitfalls That Can Kill Productivity
Are you stressed? Whether you're a founder, salesperson, dev, marketer or intern at a startup, chances are you've said yes to this question at some point. But stress isn't all bad.
The Bowery Capital family ended our CTO Summit yesterday with a different kind of talk from our friend Gagan Saksena. Gagan is the CTO of Martini Media, a network for wealthy influencers online. He was previously the founder of startup Ideascopic, and has held product leadership roles at Google, AOL and Netscape. Gagan spoke to us about how to reduce stress in our lives to improve performance, drawing upon his experience leading technical teams. One of the most interesting takeaways was that we should stop thinking about stress as something to be eliminated; it can be a motivating force too. Per Gagan's overall theory (shared and posited by many including Stanford's Kelly McGonigal) there's an optimal level of stress that both keeps the fire under us lit while keeping productivity high. Think of the classic Laffer Curve from Econ 101; just replace the Y axis with Productivity and the X axis with Stress Level.
So what productivity killers or relevant to you? In Gagan’s view, there are a number that relate to technology. Here, I tried to paint a more generalist view of startup stress pitfalls, paraphrasing various elements of his talk:
1) Inability to Detach
Detachment can mean a 5-minute break to walk outside. It can also mean something along the lines of Google’s 20% rule (where that time is spent on personal-interest projects). Research shows that there is an optimal level of detachment. Sometimes, overall productivity will be increased by a break!
2) Ill-Invested Time
More productivity means more time means an improved ability to balance work and life. The more contingency planning you do, the less surprised you will be about things in your life; and that includes the most minor work task up to planning en entire week. Gagan spent nearly an hour with his team teaching them keyboard shortcuts, and they hated it. But after a few sessions, the entire time became noticeably more productive.
3) Ignoring Long-Term Stress As A Decision Factor
Should I build this or buy it? Should I rent it or own it? These are impactful from both finance and time perspectives. But stress is a very real third consideration here. An overly complex / thorough approach may make you happy and “satisfied,” but it may lead to a lot of stress in maintaining it, or in the time it takes away from other projects.
4) Incomplete Utility Theory
Utility is an economic theory, but it boils down to usefulness. The concept, as Gagan illustrates it, is that for example, a $5 cup of SBUX coffee yields you more than $5 worth of usefulness (enjoyment, happiness, sustenance, etc.). But research shows we’re not always rational. And happiness is inversely proportional to stress. So we often optimize our decisions for utility without thinking about stress.
5) Don’t Measure Everything—Measure Properly
Measuring itself can be a source of stress. KPIs and OKRs are very useful. But you need to be prepared to pivot these metrics. Ask: are we measuring the right stuff? Cases of companies or teams measuring worthless things are commonplace. First, just measure what you can control (e.g. you can’t affect macroeconomic indicators. Second, understand your measure deeply before committing to them (e.g. aggregated statistics may have many root factors, some of which may not be impacted by efforts). Third, don’t suffer from testing biases (e.g. number of questions answered by a customer success rep, which should go down if he and others on the team are doing really well).
To finish off here are my quick 6 takeaways from Gagan’s talk. Whether a PM, dev, salesperson, CEO or marketer, I think you might find some of these useful in taming unproductive stress.
Wake up 30 minutes earlier
Validate approaches to decision-making before starting on them
Take breaks!
Check your email only so often (say every 15 or 30 minutes)
Most importantly: enjoy what you do; do what you enjoy
Thanks again Gagan for joining us at the Summit and (hopefully) keeping us at our optimal level of stress and productivity! And don't forget to check out @vreeland's thoughts on the same topic at the Bowery Blog here.
DeepMind: What Is It & Why Did Google Just Buy It For $400MM?
Re/code broke the story Sunday that Google is acquiring “secretive artificial intelligence company” DeepMind for an estimated $400m.
As posted Monday morning on the Bowery Capital Blog.
As reported in the article, the company reveals little else aside from that it is “building learning algorithms for simulations, eCommerce and games.” Clearly, given that the verticals here seem more like suggested applications than an M.O., DeepMind’s differentiator is a fundamentally technological one (that Larry Page likes). As an entrepreneur friend of mine pointed out, the founder, Demis Hassabis, has a PhD in Cognitive Neuroscience, which isn’t necessarily where the bulk of mainstream Machine Learning (ML) happens today. But, my friend also notes, Hassabis’ CMU blog suggests that he is attempting to merge concepts from his original field of neuroscience into new ML models. Taking a spin through his Google Scholar hits, we can see where DeepMind might be headed. Two examples in particular struck me as particularly (potentially) applicable:
Tracking the Emergence of Conceptual Knowledge during Human Decision Making (2009) Concepts lie at the very heart of intelligence, providing organizing principles with which to comprehend the world. Surprisingly little, however, is understood about how we acquire and deploy concepts … These findings provide compelling evidence that the hippocampus supports conceptual learning through the networking of discrete memories … Our study offers neurobiological insights into the remarkable capacity of humans to discover the conceptual structure of related experiences and use this knowledge to solve exacting decision problems.
Imagine All the People: How the Brain Creates and Uses Personality Models to Predict Behavior (2013) The behaviors of other people are often central to envisioning the future. The ability to accurately predict the thoughts and actions of others is essential for successful social interactions, with far-reaching consequences. Despite its importance, little is known about how the brain represents people in order to predict behavior… Our results suggest that distinct regions code for personality traits, and that the brain combines these traits to represent individuals. The brain then uses this “personality model” to predict the behavior of others in novel situations.
So it would seem like Google has just acquired a company that is making a bid at reinventing ML and computer program “intelligence” as we know it. If I had to venture a guess on their approach: by mirroring the way the brain stores information (a concept-based method, perhaps) and using that information to make decisions (prediction leveraging behavioral models of individuals, maybe). Of course, this is just me composing a best guess in lieu of more public information, so I could be off the mark (and if you have any more info, I’d love to hear about).
Either way, this is an exciting and characteristically bold area for Google to be entering; DeepMind has just been given access to the largest store of training data available. I look forward to hearing more from them. In the meantime, I realized that this begs another question, which I will look to answer with a follow-up post: what is the state of predictive analytics and ML today? Where is AI headed in the next 5 years?
The Joys & Dangers Of The “Mega Client”
IPO-bound Globoforce's $50MM customer & what startups should know about early big-ticket enterprise contracts.
As posted this morning on the Bowery Capital Blog.
Globoforce, a Human Capital Management (HCM) vendor focused on talent management, filed for a $75MM IPO last week. Their products enable the recognition and motivation of outperforming employees across an organization (hence the cheeky ticker choice: “THNX”). Per its S-1, the company is currently at a run-rate of about $170MM in revenue and ($7MM) in bottom-line net income. Another big win for the HR tech space, which has seen a spate of huge deals over the past several years, including but not limited to: Cornerstone OnDemand, Taleo, Kenexa, SuccessFactors, Workday, and another quite recent IPO, Benefitfocus. Globoforce’s success in particular speaks to the importance of employee satisfaction and growth as a counter to massive turnover risk that often represents a multi-billion-dollar hidden cost to big enterprises.
From the company’s home base in Dublin, Globoforce has built an impressive roster of blue chip clients across verticals and geographies. As of June 2013, the company’s “client base consisted of more than 80 companies, with more than 1.7 million users located in more than 140 countries using our solution in 25 languages and dialects” per management. JetBlue, for example, uses Globoforce’s “social recognition” software (SaaS + mobile) across 15k crew members and achieved an 88% boost in employee satisfaction in just three months. Other clients include Abbott Labs, CitiCorp, Intuit, Symantec, and InterContinental Hotels. Here's some more product color (hope Globoforce doesn't mind me pulling this from their website, which has a number of interesting white papers, by the way):
There’s one client in particular I want to emphasize: General Electric. Per the company’s S-1, Globoforce management believes that they “have deployed one of the largest and most geographically-dispersed single SaaS implementations for…General Electric, that spans more than 300,000 employees.” What’s astounding is that GE accounts for approximately a third of Globoforce’s total top-line revenue. As of June 30, 2013, the figure was 32%, implying that GE and its affiliates pay the company well over $50MM a year for the use of its platform. GE isn’t their only “mega client.” Another 39% of annual revenue (extrapolating from 1H 2013 numbers) comes from customers #2 – 10. “For…the six months ending June 30, 2013, our ten largest clients by revenue represented approximately…71%…of our total revenue.”
From one perspective, it’s encouraging to see that Globoforce can win and maintain such staggering contracts. It speaks to their sales and support organizations. Aside from the fact that they can absorb larger contracts, clients willing to pay tens of millions for a product (once onboard) will be willing to stick with you for a cycle. You don’t get the green light to set up across an organization as massive as GE without the proper testing / vetting processes; the huge potential switching costs alone can represent a barrier to entry (at least temporarily).
On the other hand, when losing one client could set your company back two years in terms of revenue scale, there’s a clear inherent risk. Not only would such a loss cripple a public company, but, there’s also a less obvious danger (one even more important to startups); being pressured by mega clients to the extent that product roadmaps cater to their needs. In this situation, it’s easy to lose sight of the larger marketplace. Before you know it, you may find yourself pitching a tool that looks more like something built in-house at your mega client. And that may be exactly what they want, if they haven’t already asked for contractual exclusivity. Another major mega-client consideration is the costs of supporting such large, needy customers. These costs can be especially dangerous if there are critical service elements of your business that scale linearly, like account management teams.
So good luck to Globoforce as it nears its big day. Assuming they price successfully, it will be interesting to see if we can learn a thing or two about how to win and retain some serious mega clients. I suppose it’s a core competency for a company so predicated on satisfaction that it chose the name “THNX.”
Watson & The Impending HCM Revolution
Why companies' biggest unseen cost-saving & revenue opportunities are the people they already employ.
As posted yesterday on the Bowery Capital Blog.
In thinking about the HR software ecosystem, I like “Human Capital Management” over the more traditional “Human Resources Management.” I think HCM may have originated from an attempt to find more delicate buzzword for workforce oversight. But I think the word “capital” speaks quite well to how employees should be viewed within an organization. While one typically expends a resource, capital is strategically deployed in order to grow its value over time. But today, the process is haphazard, and that’s costing businesses billions. Going forward, disruptive HCM solutions will use emerging data technologies (e.g. predictive data modeling, unstructured & public data capture, social network analysis) to help CXOs determine the true value of their “human capital” and opportunities to maximize their ROI on that capital.
Recruiting—the top of the HR funnel—has been a focus area for innovation over the last 5 years, with LinkedIn capturing the most notable new share by leveraging their powerful social network. Bersin, a subsidiary of Deloitte, estimates that direct investment in talent search and recruitment was $72B in the past year for US companies alone. Yet companies spend comparatively little on making sure those employees are happy and on-track once hired. The same firm estimates that global investment in integrated talent management systems will only be $4B comparatively. While the level of investment in talent management is increasing, growing 22% over the prior year, the disparity is still large. Especially when you consider that the nearly 13% of US workers that are “actively disengaged” at their jobs cost companies between $450-550B annually in lost productivity (per a recent Gallup poll).
Over the next 5 years, I believe that the application of more sophisticated predictive data models to a wealth of new unstructured & social data will revolutionize talent and performance management. The opportunity is just as large as recruiting, and almost definitely more impactful in the near term. Imagine not just tracking pay or seniority but learning how & when to preempt an at-risk top-performing engineer with a promotion, or getting a monthly list of high-potential mangers to fast-track. The opportunity is enormous in both revenue upside and potential cost savings.
Let’s put that in perspective with an example: IBM and its wholly owned subsidiaries employ 434,246 people. While I don’t have the data on hand, let’s just make the (probably overly) conservative assumption that the average salary at IBM is $50K per year. On top of that, let’s take 2 more data points: a 5% annual separation rate (the July 2013 12-month trailing average for business services from the Labor Bureau) and an estimate of the financial impact of employee turnover range from PwC: 30% - 250% of salary. We arrive at a total turnover cost exposure for IBM of greater than $2B annually.
Details will vary company to company, but once you factor in lost productivity from the unfilled position, hiring, potential relocation costs, onboarding & training, productivity ramps, risks that the replacement employee is inferior, etc., I don’t think this is an overstatement. IBM’s employees are the people that put to work $6B in annual R&D spend and do the legwork behind the company’s 5-10K patents filed every year. It’s clear how failure to stem turnover in high-value positions can be catastrophic.
Amongst other incredible technologies, teams at IBM developed Watson, the closest thing we’ve got to AI. Watson spends his time helping make utilization management decisions for lung cancer treatment plans at Sloan-Kettering, while obliterating 75-time Jeopardy! Champion Ken Jennings on the side. In 5 years or so, don’t be surprised if Watson (or software of his ilk) is also responsible for green-lighting your holiday bonus.
The Big Opportunity In Events Marketing
Go big, leverage existing CRM data & attribute sales.
As posted on the BoweryCap Blog earlier, with slight edits.
“Every Tech Company Wants A Dreamforce.” This is the gist of many conversations with gurus at Salesforce, AWS, Rackspace, etc. who are guiding the future of events marketing. Like Benioff before them, forward-looking tech CMOs see the benefit of event production: it’s brand, customer service, and–probably most importantly–leads. But Dreamforce and the like are insanely complicated affairs. Agencies such as UBM, and tools provided by Cvent are benefitting from the demand for high value, yet complex events and offering their services to help run them successfully.
The true opportunity, however, is helping these firms better understand the ROI impact of such events. They need to know: how and why these leads convert? Or are events better as lead-nurturing tactics? Hubspot and Marketo are pushing the boundaries of quantifying the marketing funnel, but events are still dark and hard-to-attribute. As an ex-SFDC exec put it: “we’re spending $10m on this event; why not $8m or $12m?” To answer that question, CMOs need a next-gen platform that addresses these key points:
Know Your Weight Class: The Total Addressable Market (TAM) is somewhat constrained to companies of a certain scale. Many companies simply don’t have the budget to invest in events / event marketing such that they’d consider paying for much more than a low-level tool focused on planning. The long-tail can use tools like Splash or existing CRMs to handle the logistical issues, but higher-level functions like auto-managing suppliers or analytics simply aren’t a mandate given that events isn’t their core marketing channel. In addition, many companies of mid scale just don’t care much about events due to precedent in the industry, whether or not that’s a somewhat myopic or misguided view (tech / media seem to be vastly more active than others).
Don’t Ignore Historical Data: Even for the big spenders / believers (ex. SFDC, AWS), actually getting to a reliable ROI is incredibly difficult. SFDC, for example, recognized that in order to do so, they had to collect years of event data and tie it closely (sometimes in a manual way) to CRM data. Even then, big attribution problems still exist (ex. they same person attended an event, viewed a webinar, and talked with a salesperson 4 times… what % did the event actually contribute to the sale?). This raises a big sales issue: if the core (or at least future) value prop is turning event marketing into an ROI-measurable channel (ex. proving profitability or not), and you can’t prove it one way or another, how do you convince them an extra cost for a tool on top is worthwhile? Keeping integrations into existing CRMs in mind from day one may be a start.
Prove To CMOs They’re Missing Something: For most CMOs, events are simply not top-of-mind. SFDC, AWS, Rackspace (and other large / fast-growing tech companies like Box) are way ahead of the curve. But even for the largest players amongst them, many other emerging solutions are more important (ex. web analytics, email marketing, CMS, etc.). What’s the best way to maximize attention? Incorporate some proof of lost opportunity into your pitch / trial / test period… maybe perform some free analysis of data off the bat to show events-driven sales lift. Surprise money on the table is a quick priority in any book.