What you could learn from ‘The Hard Truth About Innovation’ HBR Jan/Feb19
Rivalry whets our appetite for risk
Leaders can icnrease or decrease feeling of rivaly. The authoers aruge that when you want bold moves, you should increase the feeling of rivalry (e.g., pitting evenly matched employees against eachother). When you want to avoid mistakes, leaders should see to dampen rivalry.
Align you team values
Align the company around your vision. Survey Monkey decided to make
- celebrate good curiosity – reward question of the week, praise good questions in meetings or slack conversations
allowpeople to ask questions and be candid
- hold Q&Q sessions with teams
- hold creative sessions (hackathons) not just in your core business
CEO of Ford says that ‘corporate fitness’ is central to being competitive. Business is a dynamic environment, so you need to train to be fit for the changing landscape. You cannot just cut costs, they come back – you need to change the design of the system. The key to keeping fit is to keep culling complexity, as a forest fire does with the forest underbrush. Apply the fitness model to your products – ‘are they fit for customers’, if so understand why, if not, change them.
Creativity can be messy and it needs discipline and management. Successfully innovative cultures share the following:
- intolerance of incompetence
- rigerious discipline
- brutal candour
- high level of individual accountability
- strong leadership
Tolerance for failure but no tolerance for incompetence. Leadership needs to articulate what failures were productive and unproductive. Productive yield valuable information relative to their cost – do not celebrate failure, celebrate learning. Forgive mistakes, but make sure this does not slide into permissiveness.
Willingness to experiment but h
Psychological safe but brutally candid. A candid organisation will outperform a the nice one every time. People need to feel that they can speak openly and challenge ideas, especially ideas from leaders. But, people need help so that challenges are concise, fact based and logical. Leaders need to demand feedback and criticism, praise people who do this well, and educate people,
Collaboration but with individual accountability. Do not design and build by committee, collaboration is not consensus. Seek ideas and input broadly, but the people doing the work should know the most about the customer, company and product, and they are in the best position to make decisions when stakeholders disagree. As a leader, you need to demonstrate public accountability, state clear aims and objectives and hold yourself and the company to them.
Flat but strong leadership. In a culturally flat organisation, people are given wide latitude to make decisions, take actions and voice their opinions. These characteristics, allows flat organisations to respond quickly and have a greater diversity of ideas than hierarchal organisations. Flat
A clear brand helps people inside and outside the company. A clear brand provides purpose, direction, enhances products and makes recruiting and retention easier.
The authors argue that their brand identity matrix, can help by guding leaders through an exploration of thier mission, culture, competencies and values.
The authors advise a series of workshops to establish each of the nine outer boxes, before you finally define the central box – brand core.
Five basic network properties shape a platform’s scalability, profitability, and ultimately the sustainability.
Network clustering. The structure of a network influence a platforms ability to scale. Higher fragmented networks are more vulnerable to challenge, e.g., a car sharing app can challenge an incumbent in one city. To limit fragmentation you must make the network more dense and integrated e.g., increasing brand awaress, rapid expansion and first mover advantage can all limit competition.
Risk of disintermediation. Platforms are vulnerable if it is easy for members to bypass a hub for repeat work, e.g., cleaners, home carers, builders. Disintermediation can be reduced with strong contractual clauses, or limiting access until a transaction is complete (e.g,, no revealing an Air BnB address until the member has paid).
Vulnerability to multi-homing. If a member can place goods and services on multiple platforms, platforms risk greater competition. Uber and Lyft drivers often check rates and wait times on both platforms.
Network Bridging. Platforms that connect multiple networks have less chance of being disrupted. If users can use your platform to reduce time and effort to get additional and relevant services, then you much less likely to lose those customers.
Data scientist are often miss-used and underused. Data scientists need good data and a good hypothesis before they can really add value. Then data scientists must communicate their ideas in a compelling way to people who can help them put their solutions into practice.
The authors found that the top list of barriers to work for data scientist are:
- lack of managerial and financial support
- lack of clear questions to answer
- the results not used by decision makers
- explaining data science to others
Advice on building a better data scient operation:
- Define talents, not team members. Be specific about the capability you need, often; project management, data wrangling, data analysis, subject expertise, design, storytelling.
- Hire your talent portfolio. Hire a mix of talents – there are no unicorns.
- Expose team members to talents they don’t have. People want to grow, give them the opportunity by exposing them to new challenges and situations (with training, guidance and supports).
- Structure projects around talents. Work out what talents a problem needs, and the use this to find the right people to fix it.
Advice on helping data science team:
- Assign a single, empowered stakeholder. Have someone in charge of data science, give them real power and a seat at the top table.
- Assign leading talent and support talent. Put important business leaders in charge of important data science projects, or they will never get the focus and attention that they need.
- Colocate. Good data science is as much dialogue as analysis, being together makes discussion richer, faster and more productive.
- Make it a real team. Give the teams the people and tools to get the job done completely, and not handed off to another team to execute.
- Re-use and templates. Once you have an approach that works, write a playbook and use it again and again (refining and
improveingit along the way)