I have stepped down as CIO at Strands and accepted to be member of the Board of Directors to help the company from the strategic side. In September I started as Professor at ESADE Business School.
I personally would have never guessed 14 years ago that we’d have come so far, although we were dreaming really BIG! As we all know, all the good things come to an end, and I decided to accept a faculty position at ESADE Business School where I expect to share my experiences as technologist and entrepreneur. I am very excited about this new opportunity! But now it is time to take a retrospective look to a few useful things I’ve learned at Strands…
I met Cisko!
I met Francisco J. Martin (aka Cisko, aka aficionado) in Rhode Island (US) in 1997. We were attending the National Conference on AI (AAAI), our first visit to US, and my first time presenting a paper in a conference. For me, it was a life-changing conference. I met Prof. Ramón López de Mántaras who led and leads the IIIA-CSIC and was one of the early pioneers of AI in Europe. Actually, he later became the very first European receiving the prestigious AAAI Robert S. Engelmore Award. During the conference we also met a young student named Juan M. Corchado who is now a Professor at the Universidad de Salamanca. At that time, without Twitter, Facebook nor Linkedin (nor mobile phones!) the three of us socialized through an improvised trip to Boston with a rental SUV. That was a magic road trip that changed our lives and 20 years later we are still talking about that experience. We were PhD students genuinely interested in science and AI and pursuing an academic / research career. However we had great discussions about how to make research more applied and getting it closer to real world applications. Only a couple of years after, Francisco founded iSOCO in Barcelona (the very fist spin-off of the Spanish Council for Scientific Research), and I cofounded Iconomic Systems in Switzerland together with Prof. Boi Faltings, Prof. Pearl Pu, and Dr. Christian Frei. Iconomic Systems was successfully sold to i:FAO AG who became Cytric with Amadeus as a major shareholder.
In the fall of 2003, I sent an email to Francisco since I was looking for new opportunities. As always, Francisco very quickly seduced me about a new idea: to build a company with the mission of developing a new generation of recommenders systems based on human implicit behavior rather than merely in explicit ratings. The core idea was to generate music recommendations using the feedback on the actual music people play rather than using explicit ratings like Amazon. In that sense, the first tagline of the company was “what you play counts”. We had amazing collaborations on the scientific side to build a new kind recommendation systems including Prof. Tom Dietterich, Prof. Enric Plaza, Prof. Ramón López de Mántaras, and Prof. Boi Faltings. Based on all the innovations we generated an outstanding patent portfolio that was later on acquired by Apple, Inc.
Jon Herlocker and his student Matt McLaughlin were also part of the initial team. Their contribution to the company was key since Jon was one of the pioneers in developing recommender systems. He worked under Professor John Riedl who later became one of our strategic advisors. He had co-founded NetPerceptions, one of the very first companies bringing recommender systems to the market in the 90s.
Jon Herlocker preferred to stay in the academia though so Francisco stepped up as CEO and in only a few months of intense and exhausting work we launched MusicStrands.com: a service that provided music recommendations based on users collected feedback from music players (an article from New York Times and an article from Science Daily). Gabriel Aldamiz-echevarría was also an early team member and had a huge impact in launching MusicStrands spreading the product worldwide. As I explain later on in this post, few years later he created another company (Chicisimo.com) to apply the same concepts to the fashion industry. He is a great enterpreneur and a better person!
From the investment point of view, the first years we were supported by several investment rounds of family and friends, and the investment from Debaeque Venture Capital managed by Marc Ferrero and Patrick Raibaut.
MusicStrands dit not turn into a great business because a number of reasons. We found more promising avenues of revenue for our recommendation technology (i.e., applying it to personal finance). However, the main reason could be that we were operating by the book and followed all the rules imposed by the record labels. While our competitors Last.fm and Pandora offered full songs and risked legal confrontation, we just offered legally-licensed, 30-second clips of the songs that we recommended or automatically built playlists for the songs that you already owned in your music library. Our competitors did great (business-wise speaking). Last.fm was acquired for $280M and Pandora went IPO.
I’d like to relate a small anecdote to illustrate the potential of what were doing: in July 2007 I received an email from Daniel Ek (Spotify founder and CEO) and Andreas Ehn (Spotify first employee and CTO) explaining they had a company for a year with few engineers in Stockholm. We met in a small hotel in Barcelona and they told me they were looking for a music recommender… unfortunately, we were not able to see their potential and we declined their offer! Life goes on!
At least, we created a new kind recommendation algorithms and produced more than 30 patents that was later acquired by Apple, Inc. in 2011. We also developed one of the coolest social players ever capable of personalizing your music. It won a few international awards from Orange, Nokia, and Blackberry. Without a shadow of a doubt, the MusicStrands experience endured us as better technologists and entrepreneurs.
Here you have a fun video introducing many of the people contributing at that stage of the company!
In 2007 the company changed a lot through two investment rounds from Dalbergia and BBVA resulting in a total investment of $55M including the first investment round from Deabeque. The company changed its name from MusicStrands to MyStrands and then again to Strands. We managed to purchase the Strands.com domain from a hairdresser in LA. The shorter the name the more things the company was doing. Then, we enlarged the focus to include recommendation services in different domains such as music, people, news, videos, and so on. We start applying the same recommendation algorithms to other content beyond music like videos and packaged the technology to create a recommender that could be used as a service.
MusicStrands technology was used to create different products around the core music recommender. One of them was called PartyStrands. It was about creating a mobile app allowing users to express their music preferences when going to a bar. In this way, the system created a playlist according to the aggregation of musical profiles of the people in the pub. The overall result is that the pub was playing music according to the profiles of the customers (some pictures here!). Trevor Legwinski, former Director of Product Marketing for partyStrands discusses partyStrands at ABC News. We even teamed-up with MTV to create an interactive Times Square session for New Year’s Eve in 2007.
Strands Fitness (aka Running Strands)
Everybody who knows Francisco Martin can’t negate his contagious enthusiasm and passion for everything he does. But one of the things that impresses me the most is his incredible capability to foresee the future. At the end of 2008, after running the Portland Marathon, he envisioned one of the most fascinating business opportunities that Strands ever had. Nobody could see it at that point in time. With the advent of new GPS devices (including the one provided by iPhones) and the manifested and old tradition of runners to log their trainings in paper, there was a unique opportunity to build a digital training log that helped athletes not only to record their trainings but also share them with others. There were a multitude of innovative services to create around all the data collected (e.g., route matching, leaderboard for routes, race time prediction, injury prediction, when to buy new shoes, etc.).
Another incredible facet of Francisco is that he can push a team to build something from scratch in a matter of months. Soon in hardly two months after he ran the marathon, we had a first version of running Strands and as well as a full team of professional athletes both in the US and Spain ready to promote our application. Francisco even managed to get George Hirsch as an advisor. A year and half later Strands had 300k users and was becoming the leading digital training log around the world.
In Spain, Strands Team was composed of Abel Anton, Martin Fiz, and young folks like Dani Mateo or Sebastian Martos that are now part of the elite.
We even developed technology to automatically classify pictures using bib numbers. Yaroslav Bulatov worked on that project. It looks like what he learned was handy as he later joined Google and worked in a similar project to identify street numbers for Google maps. He’s one of Tensorflow contributors and now works for Open AI. Other team members of Strands Fitness ended up working for Garmin or Adidas in similar projects.
It’s well known that the single biggest reason why products succeed is timing. Francisco always explains that succeeding in tech startups is the equivalent to find and surf the right wave with the right board. The wave is the market, the board is your product, and your surfing skills are you and your team. For Strands Fitness, Strands found the right wave, built the right board, and had the right team. However, some times the short-term mentality of business can kill huge opportunities like that. The business unit we built for personal finance management started generating great revenues and the company completely switched the focus to it. Over the years, we had to digest how competitors at that point in time were acquired one after another: MapMyFitness ($150M), Endomondo ($85M), MyFitnessPal ($475M), Runtastic ($240M), Runkeeper (Undisclosed). Well, “what doesn’t kill you makes you stronger…”!
The first Fintech company in Europe
In 2007, Strands started a very ambitious project with BBVA in Spain to implement a PFM (Personal Finance Management) application for their online banking channel. It was the first PFM implementation in Europe! After that project, named BBVA Tú Cuentas, the company productized that implementation and became a Fintech company for banks. The project last for a year, and was supported by Francisco Gonzalez (Executive Chairman of BBVA), José I. Goirigolzarri (former President and COO of BBVA) and Manolo Castro (former Group Chief Risk Officer of BBVA). Enrique Gonzalez and David Gracia were leading the project at the BBVA side. It was an amazing experience involving many people and creating a cutting-edge innovative product for the bank. We were implementing the project for their web-based platform, and mobile platforms such as Blackberry, Symbian and iOS. After 10 years of its implementation, it is amazing to realize that the functionalities implemented in BBVA Tú Cuentas mean still nowadays the most advanced PFM in the market:
- Bank accounts aggregation
- Non-financial accounts aggregation
- Personalised recommendations on financial and non-financial products
- Financial tips advisory
- Community comparison
- Personalised and configurable dashboard
- Widgetised platform where users could configure their own dashboards!
Watch the video illustrating BBVA Tú Cuentas features.
Since then, Strands has been working for top banks worldwide including ING (Netherlands), Deutsche Bank (Germany), Postfinance (Switzerland), Bank of Montreal (Canada), Barclays (UK), TSB (UK), Abu Dhabi Commercial Bank (United Arab Emirates), Commercial Bank of Africa (Kenya), BCV (Switzerland), BNP Paribas (France), Israel Discount Bank (Israel), Nordea (Norway), Leumi (Israel), Davivienda (Colombia), Piraeus Bank (Greece), just to mention a few.
MoneyStrands.com is a website for the US market in which people can register and use our Personal Finance Management (PFM) product. It is available for iOS and Android. In the past, we have got a lot of attention in the media as we were one of the first companies offering such product. MoneyStrands got the Webby award for the Best Banking Website. Here you can watch and listen to some interviews on Spanish media: Barcelona TV, TVE, TV3, El Mundo, La Vanguardia, Cinco Días, Com Ràdio, Punto Radio and Radio Aragón.
Our Best Success: the Strands Ecosystem
One of the things we are more proud of Strands is the ecosystem we have created that led to different startup companies. Creating a technology company makes people work very hard together in passioned environments where everybody gets into the entrepreneurship mindset. That is probably our best success at Strands. These are the startup companies that emerged from people working at Strands:
- In 2011, Francisco J. Martin, Tom Dietterich and Adam Ashenfelter founded BigML with the mission of making Machine Learning easy and beautiful for everyone. Nowadays, BigML is being used by more than 80,000 users and universities around the world. Around 20 former Strands employees are currently part of the BigML team.
- Former VP of Marketing at Strands, Gabriel Aldamiz-echevarria, and his wife Maria Arenaza founded in 2010 Chicisimo, a company using Machine Learning to automate outfit advise. Chicisimo is a great example of how to build a tech startup focussed on the consumer space. Their perseverance and their skill to design the product based on collected users data is an example to any entrepreneur in the consumer space. Good luck, you deserve it!
- Jon Herlocker and Matt McLaughlin left Strands in 2005 and started a very successful career as entrepreneurs building different ventures: Smart Desktop, Inc., and Tignis, Inc.
- Two brilliant engineers, Ivan de Prado and Pere Ferrera from Strands started Datasalt, a company offering assistance in integrating and developing Big Data and Cloud Computing solutions.
- Another group of Strands engineers started Blueknow, a company on recommendation technologies for retailers.
- The former CTO of Strands, Nacho Lafuente, created Datumize focussing on managing dark data for companies.
- Ivan Tarradellas and other engineers from Strands started Appventures Games to develop games around tourism activities.
- In US, Flywire (a company with a huge business potential) has its tech leadership from Strands.
- Peyman Faratin founded Robust Links, a company formed by a group of computer scientists with specialization in AI, Machine Learning, Natural Language Processing, Computer Vision and Game Theory.
- Marc Monguió (former mobile software engineer at Strands) founded KiteBit, and then a second startup on the recruitment space. That company merged with JobFluent and Marc became its CEO.
- The most admirable venture has been led by Frederik van den Broek who passed away in August 2015 due to a brain cancer tumor. His last wish was to devote all his energy to create an app to help people in situations like the one he was unfortunately living in. This is his speech to raise money few months before his death.
The Academic Side
On the academic side, Strands organized the Recommenders 2006 Summer School in Bilbao, Spain that became later the ACM Conference on Recommender Systems. In 2010, Xavier Amatriain and I co-chaired the 10th ACM Conference on Recommender Systems in Barcelona with more than 400 participants from the industry (Google, Netflix, Amazon, and many startups) and academia. Strands also published several papers in international conferences:
- The Big Promise of Recommender Systems at the AI Magazine (AAAI) in 2011 by Francisco J. Martin, Justin Donaldson, Adam Ashenfelter, Marc Torrens and Rick Hangartner.
- Visualizing and Exploring Personal Music Libraries at the 5th International Conference on Music Information Retrieval in 2004 by Marc Torrens, Patrick Hertzog and Josep Lluís Arcos.
- Musicstrands™: a Platform for Discovering and Exploring Music at the International Computer Music Conference in 2005 by Gunnar Holmberg and Marc Torrens.
- Banking on Innovation Through Data, chapter 60 in the book The FinTech Book: The Financial Technology Handbook for Investors, Entrepreneurs and Visionaries by Marc Torrens. Editors: Susanne Chishti and Janos Barberis by Wiley Online Library.
- Big Data Revolution in the Banking Industry , chapter in the book The Big Analytics: Leaders Collaborative Book Project by Marc Torrens. Editors: Kumar, Vishal and Joshi, Deeksha.
It’s all about PEOPLE!
Nowadays, any company is mostly about people. In this personal retrospective look at Strands, I am not mentioning individually the people that have contributed to its success because it is an impossible thing to do without being unfair to someone. I will make however two exceptions because they are truly exceptions. The first exception is Jim Shur because he perfectly represents what an extraordinary engineer can do to a startup company. Jim is the perfect engineer knowing always exactly what and how to do without requiring any supervision. You know when you assign a challenge to be solved to Jim that it will be done on time, and it will work. He has so much experience in computer science that he is always capable of finding the most efficient solution to any software requirement. It is simply not possible to find this type of engineer in Spain because we do not understand that skilled and good engineers should remain engineers and not managers. In Spain, and probably in Europe, we are obsessed in jumping to the managerial positions. So, whenever an engineer have good experience and knowledge he is always tempted to stop being a good engineer to start being a mediocre manager. Can you imagine what would happen if all the good doctors start managing hospitals and not exercising medicine? This is exactly what we do in Spain (probably in Europe as well) with good engineers. That’s what happens to some universities as well: good researches follow the managing path and we miss them as researches and have universities managed without the management and business acumen.We need to learn from the US example in which good engineers can earn more and be more admired than any C-level manager.
Some of the heroes that formed part of the initial team are still fighting for Strands after more than 10 years, thanks Albert de Reina, Nacho Ruiz, Belen Lara, Ferran Soler, Carlos Ruana and Felix Garcia!
And last but not least I want to remember Frederick who gave a lesson to all of us with his humble attitude, hard work, and devoting the last months of his life to build an app to help cancer ill people. Thanks Frederick!
There is an amazing team behind Strands that is working very hard every day to help banks with the disruption coming from new entrants in the Fintech ecosystem. I am so grateful to all Stranders for the things we have achieved together!
Francisco, thanks once more for your hard work and helping me remembering many things for this post. Without your “obsessions” neither Strands, nor my life, nor this post would be the same!