How I failed at my startup

I started my company “eguruvu” with a goal - “Enable students to better prepare for college preparatory tests”. Based on the region, the preparation can begin last year of middle school or first year of high school. The age range of the target audience is (13/14) – (17/18). The barrier was to encourage everyone think about the preparation for these assessments in a positive way without inducing psychological fear or pressure.

Below is a comprehensive list of different functional areas. The assumptions we made that led to missteps in the execution. Cumulatively all the decisions did not lead to a successful outcome.

Customers and Segments

•  The customers were students, teenagers. Though many are interested in studies, learning is not the first thing in the mind. This was a significant barrier that was ignored due to personal bias. The founders while they were growing had academics as the priority. The same thought process carried over in the initial product development phase. It impeded our ability to do effective user research and understand the needs of customers.

• “The first mistake to avoid is to design a product based on personal     preferences. If that happens, we ignore wider market-fit and alienate potential customers”. To acknowledge and realize, it took almost 6 months.  

Total Addressable Market

• The total addressable market in one region was 125K just for grade 12 students. The four years of high school we had a total size was 600K. The country we were targeting was India. It has 25 states. There was an opportunity of similar size for at least 10 states. The total market opportunity was 6million users.  Each user $50 per year subscription fee ~ 300 million USD

• The gap in these type calculations is many. First, there is an assumption that you can walk into school and sell the product without any barriers – i.e there was a preconceived notion that we are the only one who are working a type of product or service. The school management will be easily brought into this idea, and everything will be smooth sailing from day one.

• None of these turned out to be true. Instead, jumping into product development without enough market research costed us a big deal in the later stages.

Value propositions

• Students: The value propositions were to provide adaptive learning and identify the student’s strengths and weaknesses across all the topics that will be used in the assessments.

• Teachers: For teachers, the product provided an automated way to track progress and individualized learning based on the topics they need improvement. Not only it compared the students in the same school, but across various districts [aka regions]

• Parents: For parents, it provided a dedicated portal, that they can login and see how ready their kids for the upcoming assessments

• On the field our start up team did not invest enough time, resources, and effort in clearly communicating the value. The assumption was the value proposition was intuitive and did not explain or repeat. This is a classic example of cognitive bias. Based on the Wikipedia definition “A cognitive bias is a systematic of deviation or rationality in judgement. Individuals create their own “subjective” reality, from their own perception. It dictates the behavior in the world. Cognitive biases may lead to perceptual distortion and irrationality.

• In the context of the product, being a learning app and target audience were 13 – 18 years old, the direct customers did not have the financial means of buying the product. Our main target region was the Indian subcontinent. Traditionally, students in the Indian sub-continent did not work until they graduated from college, which was around 21 or 22 years. For this product, parents are decision makers and students must use the product and take feedback from their teachers. The triangulated business model is a complex one. For the product to be successful and effective, it needs to win all the partners in the triangle relationship.

• The consumers are still evolving as individuals and the teenage years are the most rebellious period. It was tricky navigating the perception that parents and teachers are recommending and there was a fear of aversion.

• From teachers’ perspective, what is in it for them? Learning of this new tool and process.  How would it simplify their life – instead of second guessing and leaving it the chance, there was a data informed way of assessing how much the students have learnt and their ability to apply in a test setting.

• Based on the channel strategy we chose there was an inherent intent the schools will shoulder some of this responsibility.

• Barrier: The cost of the initial HP Palm Pilot device was cost prohibitive.  Instead of appearing as an additional burden to the parents, we had to develop a pricing module and structure that included with monthly and school fees.

Channels

• The channels for selling the product and service were – 1. Directly to students 2. Boards/Institutions that run the schools 3. Govt agencies that run schools 4. Charter type schools. We made a conscious decision that selling to individual students and parents might be uphill battle and did not explore much in this direction. The number of schools in the charter system were limited and return on investment for the effort did not seem attractive. The direct student selling, and charter were not closed channels, but not the first one to begin with.

• The decision was to focus on private boards & institutions and government agencies that run schools. The schools in these groups were operated in very different set-ups. Basically, we had to develop two different go-to market strategies for these two channels.

• Customer Relationships

• Now that we identified the partners to roll out the product, we focused our time and energy to collaborate with the institutions. Some of the strategies we used to enable he partnership and collaboration was to

• Digitize the questions and the responses

•  Build the system to roll out assessments

• Develop different combination of tests for personalized learning

• A support system for less technical administrative staff to install, run and monitor

• Digitize the questions and responses: We worked with an army of humans and meticulously moved a bank of 100K questions across various topics.

• Partner with the institutions to customize, install and deploy the solution based on their individual needs.

• Train the teachers and operators to schedule tests, randomize questions and responses. Support in analytical generations and reading the reports. Help adjust the course content and material to make it stick

• Develop adaptive and incremental knowledge assessments to ensure the students grasp core concepts and its application

• Revenue Stream

• The revenue stream was the product will be sold to the private and state managed institutions. We would charge per student per year subscription fee. The fee will be included in the tuition. For state managed schools, the money would be paid through a contract.  

• In hindsight I realized that selling in the educational sector is hard. The contracts need a vetting period. After the contract is signed the payment terms are not convenient. For state managed schools, there is red tape and needs constant engagement and follow-ups.” The lack of experience in selling and go-to-market cost us the company. The product was rolled out for beta testing and after 60 days moved to a paid model. Due to operational challenges and lack of commitment to adopt the new system from administrative staff, the subscription model did not materialize. The company was burning cash and no sight of incoming revenue.

• Until the sales strategy is clear and has people in the field, don’t think about rolling it out. There are no opportunities to recover from these types of false starts. The cash in the hand dictates how many chances entrepreneurs get.

Activities

The main activities included finding early adopters that can provide the feedback for the product

• Hiring the staff for digitizing the content that our can work on our learning platform

• Invest in product development

• Developing the IP of automating assessment (tests) generation from the question banks

•  Develop pattern recognition for responses

• Gamification and Adaptive techniques

•  Analysis of signals for comprehensive and true learning

• Dashboard and visualizations for – students, Teachers, and parents

• All the above activities needed substantial investment.  I and the founder pooled in our own money to kick start the activities. A huge mistake was not adopting an iterative product development approach. The focus was on building the product and getting ready before the academic year kicks-in. The lack of networking with potential buyers and users was a drawback.  Few insecurities like others can steal the product idea and impeded going overboard with marketing and advertising until we had a working version. The product launches are always tricky – if the product is out early and does not meet the expectations, the not so positive first impression limits future usage and adoption. On the other hand, if launch waits longer, lose the market opportunity, or becomes irrelevant.

Partnerships

• The founders relied on partners for the expansion of the product. The initial target market was India. India has 25+ states and their local dialects. Most of the states use English as the primary language of instruction. In the beginning English was the only support language on the platform. However, the problem is relatable in the products that need localization support.

• The idea was once there is enough traction and usage to provide support to other languages based on the market size and opportunity.

• The second challenge was that each of these states did a bit different approach on instruction and assessment. We had to rely on these partners to provide the expertise from the local states and help configure the platform that suited the users from these region

• One sales side, the idea was to sell directly to boards and institutions that ran the private schools. However, to deal with state agencies, needed a different type of expertise to navigate the red tape. We had no choice but to rely on lobbyists to get connections, exposure, and opportunity to present before the authorities who made decisions.

• Establishing and driving this partnership needed considerable investment of time and different skills. We did not have the time nor the expertise. In addition, we were operating remotely from states. The only option was to hire local experts to run the partnership. Apparently, it turned out to be hard to find good partners. We ended up with a few mediocre ones. In the end, last the money and time

Cost Structure

• The bulk of the cost was on the product development and kick starting the operations. We had the expertise in the development, but time was limited. Assembled a dedicated set of 10 engineers in the local cost centers.

• Provided 3 months training for the ramp up and teaching about the development platform. We did not do the best given the budget constraints. There was a discussion of providing equity to potential staff but did not happen.

• Exposure to equity sharing was limited. Since we were investing our money, it was not an option. We floated the idea while searching for investors at the very end.

Conclusion: We closed the company after 24 months. Lost money but the experience was priceless. Two takeaways – 1) Never do it alone. To make a product business successful, a lot of things are in play. As a founder, you might not have the experience or expertise. Collaborating, partnering, and seeking feedback is the only way out.  2) Incremental roll out and iterative product development. Don’t be too attached to your ideas . Be ready to exit.