1 of 12

Failure...what we don’t talk enough about.

Maryam Mazraei, Founder, Autopsy

2 of 12

Within 3 years, �92% of startups fail.��(Startup Genome)

3 of 12

2000 company failure submissions�12k+ subscribers�100k+ pageviews in first 4 days�#1 on Hacker News & Product Hunt�Tweeted by Marc Andreessen�Tons of press coverage on...

4 of 12

5 of 12

What do we do

�We are a data and research company that focuses on collecting information of tech firms which have failed (historical data), company founder/team profiles and zombie (distressed) companies to provide insights for VCs, angels and startups.

While there are many companies that have focused on collecting data of successful firms, the resulting survivor bias creates fundamental problems for any resultant findings. We seek to overcome this by also collecting information of firms which have failed.

6 of 12

7 of 12

Strategic Factors	Description
Product Market misfit	Product and service do not find market fit because market does not exist. Includes failure to find beachhead customer, inability to cross chasm (find next customer beyond initial).
Market failure	Market exists but company fail to find it. Includes failure to expand to appropriate product category, and geography
Market mistime	Market exists but only in future.
Pivot mistime	Company changed strategic direction before market turned.
Pivot failure	Company changed strategic direction but into a wrong direction.
Lack of capital infusion	Company failed to raise enough investment.
Superior competition	Company is outcompeted by other companies.
Pricing misfit	Product pricing is not suitable for the market.
Lack focus	Company does not focus and undertook too many different initiatives

These are some of the key factors derived from reviewing failure case studies that are used to tagged causations of failure/challenges faced prior to closure per company. We tend to try to tag between 13 to 14 variances of factors per company to help with our predictive model training which runs on companies in the same sector and business model.

We then give these factors numerical value to determine a scoring mechanism for each company on how they rank within each tag, which ultimately helps our analysis.

The factors are divided into Strategic factors which includes product, market, competition and business model issues alongside Operational factors which includes team, resources, marketing and legal issues. Each failure category has 1-3 statements that reflect the underlying concept for i.e. product-market misfit - that is commonly depicted through the data.

Other than using quantitative data analysis, we are also working on utilising unsupervised machine learning and topic modelling on all the narratives and literature review derived towards a company’s downfall to work around issues of bias around self-reported data.

8 of 12

Operational Factors	Description
Team misfit	Team does not have necessary expertise for the company. Includes lack of expertise, failure to find suitable advice, etc.
Lack of resources	Lack resources to achieve operational goals
Cost	Production cost too high. Fail to develop sufficient cost efficiencies.
Marketing	Poor marketing to reach customers.
Team conflict	Team members do not exist in harmonious fashion. Includes communication failure, lack of similar vision etc
Compliance / Legal	Company ran into legal / compliance issues

9 of 12

10 of 12

Quick exercises to do with your team:

�1. Try it for rapid validation of new product ideas in workshops or product sprint�

2. Take key failure factors and ranked them bi-monthly to track performance over-time

Although, most of you in the audience may be sitting in the product team of your organisation and may not see aspects such as marketing or being legally compliant your issue. We would like to suggest taking into consideration the majority of these factors in your product build procedures to give you a more rounded perspective and approach to give your product differentiation.

Two practices I suggest from how our findings have been used within companies:

Board of Innovation case study - utilised a bespoke list of healthtech companies we curated for them, to show a list of companies and products which had previously launched in the market but failed as a form of ‘rapid validation’ for their client’s tech tool ideas. Their client was a large pharmaceutical company. �In this scenario it was the innovation managers and corporate venture client which derived value from this dataset for their product sprints.�
Taking the factors we have captured and ranking your product and business performance from 1 to 5 every 2 months to see how this has changed over-time. Ensure, you add 1 to 3 statements under each factor and see how that changes over-time. ��The point of this is that your product manager or CEO, reviews these to give it a score or performance over-time. ��Internally, we do this in-house for clients on a bi-monthly basis and give back an executive summary of pitfalls they should be looking into immediately with a failure likelihood score compared to last period. This essentially acts like your company’s internal health-check tool. Our analysis is done via structural equation modeling and logistic regression for a binary outcome, in this a factor being a failure or not.

11 of 12

Some recommendations:

�1. ‘The Entrepreneurial Myth’ by Louise Nicolson�

2. ‘Black Box Thinking’ by Matthew Syed

12 of 12

Thank you!

getautopsy.com

@getautopsy

linkedin.com/company/getautopsy

@mmazco

http://linkedin.com/in/maryam-mazraei/

maryam@getautopsy.com