_Story:
Two startups; both are competing in an emerging market. Both rely heavily on user-generated content and multi-device interfaces; a common scenario these days. Like many early-stage companies, these startups have small teams and limited resources. They also need to deal with high-level issues & limited resources. Consequently for them, focusing becomes increasingly harder, and identifying what is important becomes confusing.
Eastern cultures have a way to deal with this state of confusion. To understand this, first let’s to break it in two general elements: Time and Tasks.
Starting with Time. Richard Lewis recently published an article on How Different Cultures Understand Time.
“West sees time passing without decisions being made or actions performed as having been “wasted.” East does not see time as racing away unutilized in a linear future, but coming around again in a circle, where the same opportunities, risks and dangers will represent themselves when people are so many days, weeks or months wiser.”
The second approach resonates with you?.. startups are tough.
His thoughts on how Eastern cultures experience time is particularly interesting. I grew up in Middle East — a Multi-Active Time Culture — , founded companies during my 20s in Southeast Asia — a Cyclic Time Culture — , and now studying in the West, — a Linear Time Culture. Let’s focus on the Eastern cultures (see side comments):
East enjoys contemplation and reflection. It is because of a sense of humbleness and collectiveness that is embedded in their cultures. They spend sometimes months, to come to a decision, yet they are very dominant in the global economy. It is a very intuitive and iterative process. Intuition and iteration are the heart of a startup. They are with the founder from the creation of the idea to the day of the exit. Circular Time is in tune with these core characteristics of lean startup method. The market, however, moves according to Linear Time principles. Past has to be settled quickly, present has to be dealt with systematically, and future has to be forecasted accurately.
Let’s focus on the second element now. Tasks are the reason that we allocate and prioritize our time. In a startup, tasks can be divided into two major categories:
- Creative tasks focus on the future sustainability (strategy, vision, culture, IMC, etc.).
- Logical tasks focus on the current realities (customer service, user experience, current supply chain, etc.).
Circular Time is particularly useful in Creative tasks. Creativity and imagination is needed in defining, for example, the startup’s vision. Vision development and discovery is an iterative process. It requires self-reflection and self-correction. In contrast, Logical tasks have to be completed according to Linear Time approach. A startup cannot spend time contemplating and reflecting about a customer support request or a questionable user-generated content. Handling these responsibilities requires a structured approach. This paradox is the main source of confusion for many young startups. A young startup is energized to move towards a vision fast, and discover the opportunities along the path. Early on, chances are uncertain, and visions are unclear. The startup is stuck with many Logical tasks that it can’t reflect on the inner-workings effectively, hurting its chance for becoming sustainable. A startup needs to have plenty of Circular Time to allocate for the Creative tasks, and one way to do this, is by reducing the time that is needed for the Logical tasks.
Going back on our two startups again. Assume that their core business model depends on user-generated content — another common scenario. Therefore, an important Logical task needs to be done: monitoring user-generated content; it is a daunting task. Copyright infringement, privacy invasion, questionable submission, unrelated content… name it, actually, let’s even outsource or crowdsource it. Both startups create interfaces for their operators/users to monitor and approve/promote the content:
Each item takes some seconds to review; that is almost an hour of the company’s time. As the startup grows its user base gets larger and monitoring UGC would become increasingly more time consuming, and consequently, less time could be assigned to Creative tasks. What is the solution to achieve systematic control on any size of data without compromising resources? I will cover two examples.
Let’s start with an intuitive approach: changing how interface represents the data. Currently both companies, display each content individually, yet the reason that they monitor the content is lack of trust in users. Why not group the content by their corresponding users?
This solutions refocuses the decision-making task on user level. Now, Company A needs to spend just few extra seconds on each item. Consequently, It saves hours on Logical tasks every day. Re-visualizing data like this, usually, requires changing only a few things, yet it results in substantial savings in the startup resources over time.
This approach, however, has its drawbacks as well; overtime complexity would creep back into the interface, requiring the startup to create lengthy and detailed operation manuals. True automation goes one level deeper. It frees the user from the mechanical and repetitive tasks, while educating them how use the device more creatively.
Circular Time Cultures are particularly obsessed with this reductionist approach:
It looks crazy to us, but we somehow want this. You know the pain of sleeping in the subways. By allowing another device to perform her Logical task of not falling, she allocated more time to her Creative task of dreaming.
The question is, how we can create a system that acts like this hat for us. What we have is our data, and what we need are our decisions. Therefore, the answer is a quantitative approach that extrapolates our data and then makes accurate decisions for us. That is an Artificial Intelligence. Let’s plan one using simple behavioral data from our operators together (they are outsourced for “savings costs” anyways.)
Assume that our two startups operate video sharing apps. Users upload the videos and the operators, usually, look at the thumbnails or short previews; either approve or remove them. That is the job to be done for the operator. Simultaneously, the backend is designed to record different predefined features of each item alongside with the operator’s decision regarding the item. These features vary depending on the medium format. For a video file, they can be defined as: words in the title, duration, corresponding user, — and colors in the video, the audio patterns, arrangements of the cut scenes, etc. The startup is responsible for providing this data, and ensuring their anonymity. It can either record it or extract it from different media and behavioral sources. Many open-source applications are available for free to simplify this data-collection process. By quantifying and feeding this data to simple — or complex — probability models, those patterns that drive the operators’ decisions can be understood. By feeding those patterns to the backend, the decision-making process can be automated:
Company B, now, saves days and considerably reduces the number of the outsourced workers. The available resource must be allocated for Creative tasks or integrating automation in other repetitive tasks:
My approach:
For each two-three redundant workers (as the result of automation), hire a professional who is capable of creative thinking and representing his/her thoughts quantitatively. You want to increase the total “creative” power of your startup while reducing the “mechanical” waste using automation. AI is to data, just as mass industrialization and automation was to assembly of raw materials. It should allow you to focus on diversifying your product, and offering more personalization.
Hugo de Garis and Chinese started this approach in Asia a decade ago. Our startup cultures must embrace this now. Re-visualizing and automating small things will result in considerable savings. It allows startups to focus on Creative tasks. “Slow and steady wins the race.” Moreover, the opportunities that come from analyzing data will allow startups to provide a greater level of interactivity to their consumers. For example, while Company B was analysing the color information in the videos, it could also look into building a Semantic Search functionality, using the collected colors. Semantic Search, in return, can create better user experience on, for example, smaller devices such as Apple Watch, and better UX would lead to ___, fill the blank with any positive business outcome that is desired.
If you are interested in data-driven approach, take a look at GraphLab (started in Carnegie Mellon University in 2009). Adaptive Automation is a source of competitive advantage that has, usually, been overlooked by tech founders and innovators. Search around and bring people who can help with this. ROI of early automation can secure future creative sustainability of any lean startup — especially now, with access to Big Data.
_Idea:
Upcoming generations not only expect connectivity and sociability, but they also demand customizability and selectivity. The digital culture gaps between our generations are widening. Human operators won’t be able to bridge these cultural gaps, at least without being biased. One way to handle this upcoming issue is to develop and trust our data and probability models, and their recommendations; then allowing them drive the user’s interactions with our applications, under our supervision of course.