Dare to outsource your ideation to a machine

Forget humans… machines can give us unbiased ideas at great volume. Ideas and employee ideation have been a great element of pride and turf battle in corporations. Whose idea, what process, who selects and cultural biases have made new idea led product development processes, a low probability success activity. Ideation has been seen as a sign of corporate intelligence. Can machines use Artificial Intelligence and take over this bastion? Can AI be the new corporate intelligence?


It can, and let me illustrate through three ideation avenues (Figure 1). Ideas come in all shapes and forms. This blog will focus on innovative growth ideas (and not performance improvements). Companies have started outsourcing ideation to a machine and the examples below are actual outcomes provided to corporations.



Figure 1


Traditional ideation buckets are market backward (ex: ideas around problems, customer feedback etc) and technology forward (ex: ideas around using a new trending technology). In a startup driven world where companies and business models are getting disrupted let’s add another bucket called “disruption onward”.


Disruption onward


The class of ideas we are looking for under this bucket are aspects that can disrupt a business or a market which is typically associated with startups.


Hence, here we make an AI-enabled machine to crunch global early-stage startup data to figure out what ideas are people pursuing. You analyze the early stage because they are the closest to the idea stage.


Below (Figure 2) is an example of idea formation in Fintech (financial technologies). The machine is fed with descriptions of startups. This is converted into vectors and crunched to identify the trends (Process has been addressed in a previous blog: “De-Clutter Life” (AI to de-clutter life)). Personal finance apps, crypto exchanges, wealth investments, decentralized block-chain platforms, AI-based stock trading etc seem to be some of the big ideas people are following. One can deep dive to figure out what is the business model of these startups and is a great list of ideas to start your company's foray into Fintech.



Figure 2


Here (Figure 3) is a dive in for ideas in Insuretech for a major insurance corporation planning a new venture strategy. DIY (Do-It-Yourself) seems to be an idea theme combined with IoT enabled data collection and offering customized insurance to various life contexts like home, health, travel, commerce etc.


Figure 3


The polygon chart on the left-hand side provides additional insights as to how these various ideas overlap. (See an earlier blog for details: “Inception: AI understands Mega-Trends”). One can also generate ideas around the white-space between the polygons or dots to seed new innovations/find business gaps. The right-hand chart plots all the startup ideas in a 2-dimensional space so one can deep dive on interesting business models etc.


This is not possible in employee-driven ideation processes. The machine gives an inherently unbiased set of ideas. In addition, the ideation process is now interactive, contextual and clickable for details.


The AI machine can “disrupt” the ideation process for finding ideas that can “disrupt”.


Market Backward


The class of ideas we are seeking here revolves around an existing market or product where customers have feedback and that feedback can be the basis of innovation.


The tenets of ideation by a machine that we explored in “disruption onward” continue to hold good here — vectors, unbiased, interactive, contextual & clickable for details. The source of data of-course changes.


A great source of customer feedback is social media and eCommerce sites. Single product customer feedback can be de-cluttered and ideas around feedback can be easily generated. (Covered in an earlier blog: “AI can de-clutter crowd wisdom in online marketplaces”).


We illustrate here a way of getting ideas about a market (with multiple competing products). Let us take a simple product like dust pollution mask which people in polluted cities use it as a protection against air pollution. How do we generate ideas around this market and not a single product? Here is where AI truly gets intelligent. Every eCommerce engine has an AI-enabled recommendation engine. The machine’s algorithm makes the eCommerce recommendation algorithm recursively recommend. You suddenly have a landscape (Figure 4) of 24 dust pollution masks spanning 6 brands with a combined 5000 customers providing feedback. Viola, a competitive landscape is auto-generated. Let’s de-clutter this and you can see how each brand fares and what product features differentiate. This is the core input for ideation.



Figure 4


Being a dust pollution mask that is worn on the face the material used is key given that it can smell due to accumulated dust. The sensation, the mask leaves on the skin is a critical idea theme. Any form of irritation has to be eliminated and the above landscape will point to a competitor product that has positive reviews around that. A clear idea emerges with an example also provided of what is a benchmark for less irritation product based on customer feedback.


Technology Forward


We look for a class of ideas that are based on technologies to come and being researched right now. The features of ideation are the same while the source changes. In this illustration (Figure 5), the machine crunches research publications around Unmanned Air Vehicles (UAVs).



Figure 5


If you didn’t know that drones can be used for food and eCommerce delivery the machine just gave you that idea. The machine also gives ideas around optimizing UAV flying trajectory for a mission, managing energy trade-off between trajectory and cellular connectivity etc.


Even if you are new to an area like eCommerce which is investing heavily in drones unlike a human, it won’t look down upon your ignorance but in an unbiased way tell you drone delivery is the future… ideate.


The oft-repeated human idiom “ideas are dime a dozen”… had a boundary of a dozen… AI just made it 1000s if not millions. But these ideas you can interact with and connect dots differently. The value is different.


Let me leave you with something mind-bending on the value of such an ideation approach. Let’s say the market idea vector space is M and the Technology idea vector space is T. What is mathematically (M minus T)?


Below (Figure 6) is “Market(M) minus Technology (T)” opportunity map for “Software Defined Networking ” (a trending investment space in telecommunication industry) delivered to a major consulting giant. The opportunity map plots ideas in the context of what the world is saying… the first step to a business case.


Figure 6


The opportunity above maps out ideas against market buzz/ feedback for various categories of innovation. Example:


  1. Product-build — as a lot of market ideas are floating and so are tech ideas in research… connect using a product.

  2. Solution-ing — Digital transformation needs technology interventions which are ideas to be pursued.


Employee ideation can’t directly auto-generate opportunity maps… machines can at incredible speeds. You can change source, lens, and thresholds to generate multiple opportunity maps at great velocity for efficient decision making. An idea is more than the dime a dozen.


Note: Companies interested in leveraging the above IP in their organization/ offerings can contact via Linkedin




Contact us

Decodem Technologies Pvt Ltd

402, Building 9, Shanthi Park Apartments, Jayanagar 9th Block, Bengaluru 560069

CIN: U72900KA2019PTC128245

  • LinkedIn
  • Twitter
© 2020 by decode'm.ai. Proudly created with Wix.com