Send-off Announcement: Exiting Our Tech Startup
Keystone Learning Systems has had a great run. It’s been a tremendous journey. The time has come to say goodbye… And hello to new opportunities.
Entrepreneurs are notorious for persevering. We adapt and pivot in relentless search of the latest sweet-spot. We reframe our weaknesses as successes. We make monumental sacrifices for the good of our cause. We remain pathologically optimistic.
We are Keystone Learning Systems, and our ed tech startup is being acquired by a California-Colorado venture roll‑up. The time has come to say goodbye… as well as a profound thank-you for the chance to make our world a better place.
We hope this open letter from us to you does justice to our users, investors, advisors, directors, and others. More importantly, we hope it serves as an entrepreneurial guide, historical archive, and adventure story for startup founders working in technology and beyond. Without further ado, here are the whos, whats, hows and whys of our work all these years startup: Keystone Learning Systems.
Contents of this Letter
What is Keystone? · Humble Origins · Broadening Our Horizons: OPUS · Back to the Future: Pivots · Back to the Future: Market Research · A Phoenix Rises: Keystone Learning Systems · Seeking Legal Council · Breaching New Markets: Cocoa.io · What was Cocoa.io Anyway? · Intelligent Learning Resources Algorithms · Diversifying: More B2C Ventures · Individualized Learning Platform (ILP) · Keystone’s LEAP · Consulting for Others · Dissolution: The Implosion · CTO Ryan Atallah · CFO Ved Topkar · CEO Carson Kahn · Pivots: Food for Thought? · Goodbye: Letting the Sun Set · Into the Future: Arktos Data Systems · Stepping Down As CEO: What’s Now? · Lean Startup: Lessons Learned · Exiting: What Comes Next? · Thanks
What is Keystone?
Many of you are familiar with our company, and know we’ve been seeking acquirers. For those who aren’t, however, a quick primer:
Keystone is [was] an educational technologies startup dedicated to overcoming far-reaching problems in secondary education through creative applications of cloud computing and social technology. We see the digital frontier as a turnkey medium for engaging and empowering members of the school community to collaborate in ways worthy of the potential in every student. Our products and services nurture and extend this potential by individualizing, optimizing and modernizing the learning experience — in and out of the classroom. Inspired by our own experiences in school, the entrepreneurs driving innovation at Keystone strive to introduce fresh perspectives to the use of web tools in education. We’ve even partnered with leading experts and research institutions to hone and perfect our solutions. Tens of thousands of students, teachers, admins and parents depend on and enjoy our products. In short, we aren’t waiting for education to reach the 21st century… We’re bringing the 21st century to education today.
Our Humble Origins
Like many great startups, Keystone began as an informal collaboration among friends. In 2010, visionary Alexander Atallah, developer Ben Alpert, designer Ryan Atallah, technologist Ved Topkar, and I (Carson) were recruited to “build a new website” for one of the nation’s top-ranked secondary schools, Fairview High School (we had about forty years of design, development, and marketing experience between us). A couple years prior, I had been Product Strategist at Vetica Interactive, the software-turned-educational-technology startup later responsible for PlanboardApp.com, a popular curriculum design and lesson-planning web app for teachers (which has since become Chalk.com).
Like other Colorado-area schools at the time, Fairview High had been running Microsoft SharePoint 2004, and its web presence was lousy at best. It didn’t reflect the forward-thinking, technology-driven pedagogy that earned Fairview its high rankings, nor did it tap into the students, teachers, admins, and other key figures that made Fairview a model example of public education in the United States.
Early on, we learned to pick a small set of non-negotiable standards that matter most, and enforce them ruthlessly. We realized that for Fairview to reap the fullest benefits from this project, the school community needed to consider itself a true “stakeholder” in the planning, operation, and outcome. So, after ruling out education-inappropriate solutions like WordPress and Drupal, we involved dozens of Fairview students in our process of building a Ruby-powered Content Management System (CMS) for Fairview 100% from scratch. This gave the community a unique sense of purpose and control — not to mention learning opportunities — around their new web presence.
Following deployment, Fairview’s website usage and traffic increased by over 60%, saving tens of thousands of dollars in educator efficacy, paper use, and communication resources, as determined by a series of longitudinal analytics studies. Users of all kinds reported near-total satisfaction with the user experience. Our platform also boosted Fairview’s brand image, with greater awareness, engagement, and reach than ever before.
Not only were we pleased with the results, but the public at large was very impressed. Independent news media like The Daily Camera and Royal Banner began to take notice, and encouraged us to take our efforts even further. We wouldn’t stop at a CMS — we’d go on to design a suite of virtual learning tools, homework trackers, and social networks to augment the educational experience at Fairview and beyond. After all, good products don’t get built in a vacuum. In that light, OPUS was born.
Broadening Our Horizons: OPUS
OPUS was the first iteration of our efforts to expand, extend, and commercialize our product for a wider audience. In that vein, we began building ClassRoom™, a set of interactive learning tools tailored to the needs of high school students. ClassRoom included virtual flashcards, a quiz platform, as well as discussion and question forums personalized for every student. Fairview piloted these and other upgrades to the original CMS, and were thrilled with the results.
Starting with the right idea matters. As we quickly came to understand, online learning enables highly individualized and differentiated teaching methods that support different learning needs, particularly in cases of english-language-learners, students with disabilities, and gifted students. Persistent access to learning materials allows for greater flexibility, such that students can progress towards different goals or learn at different paces. More flexible learning styles also better accommodate travel, athletics or other time-specific commitments that can prevent students from engaging in active learning on a strict schedule.
Unfortunately, barriers to entry in the ed tech industry are numerous and high. They include oligopolies, decentralization, vicious sales cycles, pilot error, teacher time as sunk costs, vendor walls, and, of course, capital. These are certainly problems “per se”, but their biggest effect is on morale. Make no mistake: morale is very real and self-perpetuating. If we wanted to move forward with OPUS, we would have to figure out how to overcome these hurdles.
To complicate matters, project originators Ben Alpert and Alexander Atallah were moving onto greater things. Ben became a computer engineer at Carnegie Mellon University and Junior Software Developer at Khan Academy, the world’s most popular educational videos site. Likewise, Alex became a computer scientist at Stanford University and Forward-Deployed Engineer at Palantir, one of the nation’s foremost data analytics and computational intelligence companies. Joyously but nervously, cofounders Ryan Atallah, Ved Topkar and I bid them farewell, and then OPUS took an unexpected turn.
Back to the Future: Research & Pivots
The average tech startup pivots twelve times before succeeding. In the face of our leadership crisis, brand snafu, and lack of domain knowledge, we knew one of our many pivots had to happen right now.
The first step was rebranding our venture. Please, please take this to heart: a brand goes far beyond a company’s “image”. Indeed, that sense of the word is incredibly old-fashioned: starting in the 1800s, cattle ranchers used a “branding” iron to indicate which animals were theirs. So unless your startup is rather bovine in nature, you should start thinking about brands as evolving, customer-driven personas, with both tangible and intangible elements.
Rebranding was not a simple process. We didn’t just hope to change the tenor of our brand… We needed to change the very direction of our company. We needed to change not only how we talked about ourselves, but how we thought about ourselves. We needed to change not only what we produced, but how we produced it. And to do all that, we needed to know, deeply and deliberately, what our competitors were doing.
For many months, we engaged in in intense market and competitor research initiatives. We uncovered several generic takeaways that I’m comfortable revealing here:
- Educational and instructional technology is one of the world’s fastest-growing startup industries (top 7, by most accounts).
- As an institution, education in the U.S. is pretty awful.
- There are very few advanced technology companies serving secondary education. Our friends at Schoology and Edmodo are the most prominent exceptions, but they have serious shortcomings in technology, design, and pricing.
- Most of the action happens in higher ed, with billion-dollar corporations like Blackboard and Desire2Learn monopolizing college and university markets.
- Many schools are using Content Management Systems (CMSs), but they really need Social Learning Communities Management Systems (SLCMSs/LMSs). The market for true LMSs in 2015 will grow to $24.2 billion in the United States alone. Worldwide growth is far larger, with compound annual growth rates of up to 30%.
- When it comes to LMS/SLCMS companies, over two-thirds are commercial, fewer than one third are noncommercial, and there’s very little in-between. Just 32% of providers offer a free counterpart product.
- Barely over half of these companies facilitate community support for their products, and on average, their proprietary support receives a rating of just 2/5 stars from end-users.
- Fewer than 60% of ed tech providers are eligible for federal ERATE subsidies, which would allow public schools to pay less for technology upgrades. Ergo, the barriers to new tech in public schools are way too high.
- While over 90% of providers are fully FERPA- and CIPA-compliant, just 70% comply minimally with federal/international accessibility guidelines, and 23% don’t comply at all.
- Out of 166 elements known to be essential for effective Learning Management Systems, our average competitor incorporates only 85, or 51%. That’s a failing grade. (For reference, Keystone 2.0 aimed for about 137 elements, roughly 83%.)
- The web technologies our competitors use are grossly out-of-date, with a fourth of providers scoring as low as 0/10 and no one higher than 6.4/10 in terms of perceived quality and potential (Petrovich Index).
On a larger scale, there existed eight primary market segments we could’ve targeted, with eight types of pedagogically-defined eLearning products and services: Self-Paced Courseware; Digital Video, Text & Audio Reference; Test Prep & Exam Products; Collaboration-Based Learning; Social Learning; Simulation & Game-Based Learning; Cognitive Learning; and Mobile Learning. These products and services were delivered by providers in four formats: Packed Content; Custom Content Services; Software as a Service or SaaS; and Installed Tools/Technology. (For reference, OPUS was a SaaS venture.)
We also conducted thorough analyses of our competitors’ brands and user-bases. The long and short of it is, we determined we could do much better, and were already on the road to success. After intensive rebranding efforts, upgrades to our platform, and plans to design and develop five new products, Keystone finally emerged.
A Phoenix Rises: Keystone Learning Systems
From OPUS’s ashes, Keystone Learning Systems sprang to life. Ryan (left) became CTO and Product Director; Ved (center) became CFO; and I (Carson, right) became CEO. For anyone who’s listening, make most decisions by consensus, but have a CEO whose decisions are final. That said, your authority as CEO is earned, and your job is really to get out of other people’s way an enable them to do their jobs.
We also recruited several experienced paid and unpaid interns to help us produce some of the best cloud-based education applications available, including Ted Kornish (media and marketing; now a designer at Ibotta), Aaron Perley (web development and design; now a programmer at Ubooly), David Phan (research and development), and Logan Garbarini (information systems).
Of course, we assembled an incredible and indispensable Board as well. Members included (in order of signing): Dr. Francie Anhut, former CEO of Impact on Education; Mr. Matthew Piskorz, Associate at Tango Capital Investments; Dr. Jonathan Bein, Managing Partner at Real Results Marketing; Dr. Feng Wang, Director of Online Learning and Academic Technology at John Jay/CUNY; Dr. Laura Kornish, marketing guru at Leeds Business; Dr. Hillard Jason, one of NLM’s star educators; and Dr. Gary Kahn, CEO of Healthbridge Systems LLC.
Furthermore, we developed what I call an “advisor/investor connection map.” I highly recommend this practice to other startup founders. It consists of a simple, bootstrapped CRM to track instances of contact, degrees of separation, and key relationship information for current and prospective advisors and investors. It doesn’t have to be big — ours held fewer than 60 entries — but it helps founding teams and their employees keep track of networking and fundraising progress. That said, never fall in love with the fundraising process. Also, most investor advice isn’t great for building a magical product. No one can “help” you build a magical product — that’s your job.
We sought legal council from the Entrepreneurial Law Clinic (ELC), a CU Boulder initiative spearheaded by the distinguished J. Brad Bernthal of the Silicon Flatirons Center for Law, Technology and Entrepreneurship. In addition to Mr. Bernthal, who oversaw our term, a team led by the dynamic Ms. Katherine Nelson and Mr. Brian Taylor accommodated our day-to-day needs. They assisted us in sorting through over 35 claims to unique intellectual property, refreshing our partnership and vesting agreements, and matters of international internet law. If a university or accelerator near you offers a similar law clinic program, take advantage of it right away.
We also continued attending educational technology events to network around our new brand. One such event was the famed Association for the Advancement of Computing in Education World Conference on Educational Media and Technology (AACE EdMedia) in Denver two years ago. We took the crowd by storm. My advice for other startup founders is this: attend industry conferences when designing a new product — not just marketing one. As it did for us, this will expose you to what potential collaborators and competitors are doing with a depth and breadth you just can’t get from Google Alerts or trade magazines.
Breaching New Markets: Cocoa.io
At this point, our B2B platforms had empowered tens of thousands of users, and we decided it was time to diversify into B2C models. This decision resulted from an extensive consumer research survey of 100 questions covering perceptions and emotions, skills and tools, school management, and meta-feedback about:
- Online educational tools
- Trust as a decision-making factor
- Attitudes towards online learning games
- Opinions on social media
- Important qualities of software
- Individualization and personalization
- Adoption rates of ed tech
- Peer-peer communications
- School branding
Why did we use a consumer research survey? Educating a market that doesn’t want your product is usually a losing battle. Stick to your ideals and vision, but respect trends. Also, quantify everything, and qualify pathologically. In that vein, we started working on something pretty cool: Cocoa.io.™
Cocoa was the world’s first community-driven study toolbox that learnt how you learn to help you learn even better. It conformed to a branch of educational technology known in the industry as a “personalized virtual learning environment”, or PVLE — software or webware that combines machine learning technologies, social networking practices, and gamification strategies in the context of individualized learning.
We started developing Cocoa — including an advanced learning personalization technique now known as IntALRA™ (Intelligent Aggregation and Recommendation of Learning Resources Algorithm). To give you an idea of how the now-retired product functioned, here’s the midsize pitch letter we used for Cocoa:
Meet Carson, a high school student who, like 97% of high school and college students, knows he needs to study to meet his goals, make his grades, and learn the material. Like 63% of them, however, he often has trouble channeling his motivation towards studying, and when he does, he retains only 46% of what he reviews. But Carson doesn’t know that. All he knows is that he wants to minimize the bother, maximize the benefit, and be the best student he can be.
So last week, Carson started using Cocoa (cocoa.io), a fun, attractive cloud-based study toolbox that boosts, tracks, rewards and channels his motivations towards the most effective and enjoyable learning possible. Carson loves Cocoa as a modern way to do what he’s always done – study – but with much less bother, way more fun, and measurably greater success.
But behind the scenes, Cocoa’s data-driven platform blends the addicting stimulus of gamification, the productive influence of social communities, and the intelligent power of content aggregation and recommendation to create “positive feedback loops” within users, the application, and the Cocoa community alike. What does this mean for Carson, and for students everywhere? It means Cocoa becomes even more effective, fun, and social every time it’s used, which is good, because it satisfies a need that’s never going away.
Carson, like 83% of his peers, uses flashcards as his primary means of studying. Like 57% of students, Carson’s love of the internet led him to try a few online flashcard websites, but he found them no easier, no more fun and no more responsive than his trusty paper companions.
When he discovered Cocoa via his friends’ Facebook activity, however, he found it actually boosted his studying in ways unachievable with paper cards. He could track his progress and stay motivated with points and badges, turn his stacks of cards into interactive quizzes and games, study from decks made by his friends, and receive meaningful answers to questions he had about tough topics. Carson found that as he engaged with Cocoa in more ways, he could earn credits towards the purchase of wonderfully useful features, and he could compete against his friends or students anywhere to be the best he could be.
When he struggled, he received personalized recommendations of Khan Academy videos, scholarly articles and Cocoa stacks he could use to fill in the gaps and even learn something new. Cocoa minimized the bother and maximized the benefits of studying, and only with Cocoa, Carson found, could he truly be his best.
As it turns out, I’m Carson, and I truly believe this could be the story of every student with a smartphone or computer. To make it possible, I developed Cocoa with other industrious college students who felt it’s high time the study experience got an upgrade. If you were ever once a student and agree with our cause, please join us in helping students everywhere be the best they can be. Because with Cocoa, anyone can.
The Intelligent Aggregation & Recommendation of Learning Resources Algorithm (IntALRA) underlying many of Cocoa’s points of differentiation was an innovative combo of machine learning vectors and social networking mechanics that determined the relevancy of educational resources, recommended them, and tracked engagement through gamification. It combined a normalized “vanilla score” with a weighted matrix accounting for the authority indicies of raters based on stack progress, level attained, points within and without of subject, badge possession, degrees of separation from stack creators, social engagement with related materials, and pre-deduction total weight scales. The algorithm refactored the results as a “scalar index”, accounting for the relevance of ratings to the consumer based on Wilson mean score intervals and multiplied means fed through Echo-State and Kohonen neural networks, which were then reported as “balanced ratings”. These data were overlayed on keyword association and metadata tracking algorithms with a series of social and pedagogical extensions.
In all, Cocoa.io was off to a good start. But Cocoa wasn’t the only product we pursued…
Diversifying: More B2C Ventures
As all startups should, we assumed the market is efficient, and valuable ideas will be discovered by other teams nearly instantaneously. So, having evaluated a variety of additional project trajectories, we extended our long-term strategic plans to reflect emerging market data. In 2011, U.S. education expenditures rose to USD$1.3 trillion, $625B of which was allocated for K-12. Just 0.5% of this market was attributed to eLearning. That said, 5 million additional K-12 students engaged in some form of blended learning, and over 30% of this activity could be classified as “social” learning (CAGR 43%).
Keystone’s Facebook-Based Learning Management System (FBLMS) addresses the one issue a majority of school community members across all levels of education identify as supremely problematic: communication. By leveraging the stable, secure, well-funded and widely-familiar Facebook™ platform as a technical backbone for school communications and community management, the Keystone FBLMS will help to repair and enhance those aspects of content and communication that interfere most in school environments, while innovating around the popular but untapped potentials of social media.
The three-tiered social application centers on automation, ubiquity, interrelation, reliability and bottleneck mitigation as core areas of value-addition in educational (a) operations, (b) climate, and (c) image. These aspects, in tandem with our approach to cloud-nativity and Facebook integration, promote tremendous scalability at both extremes: entire districts or teachers of a single class could derive proportional benefit from the application, which, along with antispecialism and anachronism, is a major shortcoming among Keystone’s competitors.
This data-driven product lends itself to a variety of monetization strategies, from habituation-based subscription with freemium crosslinking, to a unique partnership/sponsorship/NFPC model. All approaches are capable of bearing minimum weight on shrinking institution budgets while bearing maximum social and consumer impact. Marketably but darlingly innovative, research-supported, and highly flexible, the Keystone Facebook-Based Learning Management System (FBLMS) wields significant potential as both a standalone application and as the foundation of a robust social learning platform for future expansion and productization.
Furthermore, we pursued what’s known in the industry as an Individualized Learning Platform, or ILP. The ILP was designed to improve productivity pathways in secondary schools, by broadening access, engaging students in active learning, individualizing and differentiating instruction, personalizing learning, efficiently allocating student and teacher time, increasing rates of student learning, improving assistance to disabled students, reducing school operating overhead, reducing salary costs, and realizing economies of scale. My takeaway is this: beware of chicken-and-egg products. To understand what I mean, just take a look at Keystone’s ILP:
Students who learn in different ways are disadvantaged by education homogenization. The Keystone Individualized Learning Platform (ILP) seeks to empower students and teachers to interact through differentiated instruction that is personalized for student academic needs. Individualized learning has the potential to increase both engagement and instructional effectiveness to the end of reducing dropout rates and facilitating student personal growth. Though individualized learning as a pedagogical theory has existed for centuries, in the absence of technology-assisted scaling, it has not yet been practical.
Keystone’s ILP enables individualized learning by facilitating organized differentiation of instruction on a student-by-student basis, informed by detailed learner information. Constant feedback between students, parents and teachers directs the differentiation process, which is operated through a visual user-interface, and is aided by informed pedagogical suggestions and best-practice recommendations. Learner information is collected through profiles and skill-tracking closely integrated with course materials. Progress is continually assessed through quizzes that track mastery and areas of improvement; this information is then used to design a lesson plan specific to student needs. Finally, students are empowered to make choices about their curricula to encourage personal interest, creativity in alternate pathways, and a sense of purposeful accomplishment.
Presently, no platforms exist to make individualized learning available to learner communities in a scalable and accessible format. Some programs offer isolated, pre-made individualized courses, which provide valuable proof of concept, but are not true competitors as they fail to empower teachers to design their own courses.
Individualized learning in secondary schools is best implemented with students who can make informed decisions about their education, and with progressive teachers willing to harness exploratory teaching strategies. While a global system-overhaul is not required for Keystone ILP, a paradigmatic shift towards personalization is optimal for maximizing benefit. Current progressive trends like the “flipped classroom” are ideal substrates for Keystone’s ILP, and while this may spawn user skepticism, it will also prove advantageous in marketing efforts.
Later, we planned a community-driven textbook exchange called Switcheroo, which in a series of vague ways became the still-operational Harvard Crimson Exchange. The “community-driven” aspect was a key point of differentiation: using sales to cultivate good will and community is the future of educational commodities.
We also pursued a proposal by the University of Colorado, called the Leeds Experiential Learning Reporting & Assessment Platform (LEAP), to fold real-world internships into the academic accreditation process. Keystone LEAP was best described as follows:
Keystone LEAP aims to facilitate, track, and enrich for-credit internship opportunities through the accuracy, continuity, security, simplicity, standardization, and speed of cloud technologies. As defined by CU’s own Center for Experiential Learning, experiential learning (EL) is “a process through which students develop knowledge, skills, and values from direct experiences outside the higher academic setting. It encompasses a variety of activities including internships [and] professional work experiences. Well-planned, assessed EL programs can stimulate academic inquiry by promoting interdisciplinary learning, civic engagement, career development, cultural awareness, leadership, and other professional and intellectual skills: (1) Reflection, critical analysis and synthesis; (2) Opportunities to take initiative, make decisions, and be accountable for the results; (3) Intellectual, creative, emotional, social, and physical engagement; and (4) Tailored learning experiences accompanied by natural consequences and successes”.
Students deserve for-credit internship opportunities to be competitive in the scholastic context. And though no educator or employer should have to bear undue burdens associated with practicum, excessive obligations to self-monitor often distract students from making the most of their experiences. User-centric cloud technologies can relieve these burdens, and should reduce stress by standardizing assessment procedures around internship tracking. If nothing else, experiential learning is critical to academic success.
Keystone LEAP will help the colleges fulfill or exceed Clark and White’s famous 2010 assertion that “a quality university business education program must include an experiential learning component” (American Journal of Business Education). With total focus on the end-user, LEAP promises a seamless, responsible, edifying, and gratifying practicum experience to all involved. Businesses will appreciate the web platform for keeping interns accountable, while Leeds will finally be able to award credit commensurate with work experience. Leeds will operate with more integrity by eliminating the “faux 1-credit” system for students whose internships require it, bringing it to par with prestigious competitors like UNLV Lee and CMU Tepper. CU will outshine these institutions, who conduct internship audits just 15% of the time and receive documentation just 41% of the time (NYT). Keystone LEAP can make the “impossible”, possible at last.
On top of these new products, Keystone Learning Systems staff began formally and informally consulting for other organizations. The organizations had various needs — some wanted education middleware or APIs built; others needed education-specific marketing; still others requested design thinking analyses of their existing educational technologies. Our clients fell across the board, including:
- Facebook for Education
- Impact on Education
- The Association of Educational Therapists (AET)
- John Jay College of Criminal Justice
- University of Colorado at Boulder (CU Boulder)
- New York District 6 (NYCD6)
- Scrib Coworking
- Leeds School of Business
- Trade Partnership Worldwide, LLC
- Boulder Valley School District (BVSD)
In other words, Keystone had its fingers in a lot of pies. At the time, we believed it was a good thing. Our platforms were satisfying tens of thousands of users, we were attracting new business, the world seemed better off for Keystone’s efforts. The thing is, working on the wrong thing for a month is equivalent to not showing up to work for a month at all. We found ourselves in such a position, and for the eighth time, Keystone Learning Systems took an unexpected turn.
Dissolution: The Implosion
Make no mistake: this “implosion” was nowhere near as ominous or devastating as it sounds. In fact, it was the confluence of too many new priorities and our changing geographical locations which, one by one, drew us away from a steady base.
To begin with, cofounder and CTO and Product Director Ryan Atallah received offers to study computer science and economics at Stanford University, and to work as a developer at Facebook’s world headquarters. He was well-suited for the position. Ryan commands an extensive background in object-oriented development, interface design, software architecture and filmmaking. He has been a computer science TA at Stanford. In addition to countless Keystone products, he has also developed in-house web applications for a variety of industrial clients, including environmental metrics corporation Vaisala and scientific instruments manufacturer SofTA Corp. As a filmmaker, Ryan has shot and produced an array of short films, including a nationally award-winning documentary entitled “Wilson’s League for Peace”, and has twice competed in the National Debate Circuit as one of the country’s top Lincoln-Douglas debaters.
Secondly, cofounder and CFO Ved Topkar received offers to study mathematical and computational biology at Harvard University. In addition to his talents in mathematics and science, Ved has been a longtime new media advocate specializing in information systems, mobile applications, 3D rendering and motion graphics. He wields significant administrative experience as the State of Colorado’s 2011-2012 Youth Governor and Lead Representative to the Conference on National Affairs. Ved was also a research intern at University of Colorado Department of Electrical Engineering, where he studied the propagation mechanics of electromagnetic radiation in synthetic tissue structures. Ved has since been Executive Director of Science Policy at The Triple Helix, and is now Director of Technology at the Digital Literacy Project, a nonprofit working to expand access to technology domestically and around the world.
Thirdly, I was recruited to be Deputy Director for Operations and Marketing for Online Learning at John Jay/CUNY in Manhattan, who were expanding their Massive Open Online Courseware and Distance Learning initiatives. In the face of running a startup, I wouldn’t have even considered this opportunity if it hadn’t been for the tremendous industry credibility that I knew would flow through to Keystone when I returned to it full-time. I also began receiving speaking invitations, and delivered several talks and seminars to industry professionals, including “Adapting Instructional Technologies for Special-Needs Students”, “The Evolution of Learning Management Paradigms in Higher Ed”, and “Usability Heuristics for Primary Education”. Previously, I cofounded the IDEA Media Cooperative, a multinational consulting network of creative media agencies, startups, and freelancers with members throughout the United States, Canada, Mexico, Spain, Turkey and China — and though it was relatively self-operating at that point, IDEA was picking up renewed steam. And to complicate matters, I was later accepted into Watson, the world’s first university-accredited startup accelerator for social entrepreneurs.
In short, over the past several months, we’ve been in no position to entertain new business. Things began to slip, we reeled them in, and then they began to slip again. On the whole, needless to say, I believe it was an issue of focus, drive, and timing.
Pivots: Food for Thought?
In certain ways, we’ve become a little disillusioned with ed tech. The deeper we go and more progressive we get, the more resistance we’ve encountered from major players in the industry. Of course, that’s to be expected with any industry in great need of change… It’s to be expected by the change-makers, that is. But it increased the cost of doing business—and not just in dollars—to the point of requiring changes in pace, focus, and forum. Remember, businesses are human systems, and if the people in your startup are loosing focus, your startup will as well.
Still, we haven’t been able to shake the thought that, possibly, all the technology secondary schools need already exists. That is, what if we’ve been mistakenly selling convenience — cherries on top — instead of focusing on innovation at the bottom of the pyramid (BoP)?
In considering various pivots and ways to salvage our tens of thousands of man-hours, we also realized that consumer-space startups often don’t solve grand social problems. Perhaps then, we thought, we should pivot again to become a company that leverages our existing expertise to build middleware and APIs for the B2B educational technologies that already exist! That mission, coupled with effective turnkey metrics for evaluating the success of our cross-networked technologies in schools, could be a wonderful, straightforward, low-cost solution, right?
Maybe. Maybe not. For Keystone’s founding team, the bigger issue has really been our ability to commit time and energy to making education better… without compromise. We’ve come to realize that our passions — and tolerances — for these issues are running low. Pivoting will not only add strain to our professional and personal lives, but perhaps to the industry as a whole. What’s more, we have a great many other opportunities to pursue, and we believe that other key causes deserve our talents and attention.
Ultimately, after dozens of hours of evaluating the possibilities, we’ve decided that our focus is now better leveraged elsewhere. This isn’t to say we’ve “surrendered” in any way… It isn’t for lack of trying that Keystone arrived at this point, after all, nor will we give up on making the world a better place!
Goodbye: Letting the Sun Set
All told, we’re finally allowing our startup to be acquired by a large California-Colorado venture roll-up. We’ve had a great run. We’ve learned a lot. Still, there comes a time when every founding team must say goodbye; for those of us at Keystone Learning Systems, that time has come. But it’s also the beginning of a bright new future.
Former CTO Ryan Atallah has just been appointed CEO of a new Silicon Valley tech startup, called Arktos Data Systems, aiming to democratize data visualization. He’s also propagating his love for educational technology through independent venture research into knowledge and curriculum mapping tools and algorithms. And he’s sharing his skills with the world by teaching computer science topics to other students at Stanford.
Keystone cofounder and CFO Ved Topkar, on the other hand, is moving forward with IDEA Media Cooperative, the multinational creative media network we previously launched, and is also doing groundbreaking computational epigenetics research at Harvard Medical School. He’s working on genome-scale data integration and analysis solutions, and many other incredible technologies I’ll never fully understand.
As for me, I’m stepping down as Keystone CEO at the end of Q1 2014. As a design thinker, entrepreneur and technologist, I’ve dedicated my life to a single mission: making society a positive-sum game. My startups and projects in digital media, education, art, and business have touched over half a million lives and are transforming the way change-agents learn and interact throughout the world. As for what’s next, I’ve become an Associate Partner in Mosaic International, the entrepreneurship consultancy founded by the illustrious Michael Williams, where I’ll be coaching clients from Google to Ashoka to Chick-fil-A in developing more sustainable business practices. I’m also joining forces with the Dr. Jonathan Beninson in StirUp.org, to accelerate other entrepreneurs around the globe. I’ll continue with a few other big things, too. I’m 20. We’ll see where life takes me.
Lean Startup: Lessons Learned
At Keystone, we’ve had a tremendous journey. And though we’ve already discussed some of what we’ve learned in this letter, we’d like to take a moment to make it more explicit (with the help of RethinkDB’s Slava Akhmecht, of course). Without further ado:
- Whether your startup is billed as a “social enterprise” or not, always have a triple bottom-line (3BL).
- Good products don’t get built in a vacuum. Find your field and be the best in it. If there is no field, something’s wrong.
- Starting with the right idea matters. Empirically speaking, you can only “pivot” so far.
- Most investor advice isn’t great for building a magical product. No one can “help” you build a magical product — that’s your job.
- User comes first. Product second. Brand third. No degree of marketing will help if you don’t have these priorities straight.
- Working on the wrong thing for a month is equivalent to not showing up to work for a month at all.
- Pick a small set of non-negotiable standards that matter to you most and enforce them ruthlessly.
- Do everything you can not to attach your self esteem to your startup (you’ll fail, but try anyway). Every once in a while, get away. Go hiking, visit family, play racquetball, ping pong, anything. It will make you more effective and make the people around you happier.
- Make sure it’s good enough and move on. Pick implementations that give 80% of the benefit with 20% of the work. Development speed is everything.
- Quantify everything. Qualify pathologically. Don’t guess; measure. And prioritize ruthlessly.
- Make promotional videos for your startup and each of your products or services. Video is the future of promotion.
- Reevaluate effectiveness on a regular basis. Cut things that don’t work, double down on things that do.
- Beware of chicken-and-egg products. Make sure your product provides immediate utility.
- Product sense is everything. Learn it as quickly as you can. Being good at engineering has nothing to do with being good at product management.
- Make most decisions by consensus, but have a CEO whose decisions are final. Your authority as CEO is earned. You start with a non-zero baseline. It grows if you have victories and dwindles if you don’t. Don’t try to use authority you didn’t earn.
- Never fall in love with the fundraising process.
- Avoid buzzwords 99% of the time.
- Morale is very real and self-perpetuating.
- Educating a market that doesn’t want your product is a losing battle. Stick to your ideals and vision, but respect trends.
- Split the stock between the founding team evenly. Always have a vesting schedule.
- Assume the market is efficient and valuable ideas will be discovered by other teams nearly instantaneously.
This letter being digital and all, we’ll continue adding lessons to this list as we think of them. Please check back often for updates.
So, What Comes Next?
As mentioned, Keystone is accepting an acquisition bid from a North American venture roll-up. That said, we are always interested in collaborating with new partners and hearing from interested parties. Because we have just retired our official web presences, please contact us directly to get in touch: email@example.com
We hope you have found this letter informative, cathartic, and sincere. Please don’t hesitate to leave questions or comments in the margins. We are so grateful for your support throughout the last few years, and look forward to many incredible adventures yet to come.
On behalf of everyone at Keystone, thank you.
Carson Kahn, CEO
November 26, 2013