What you could learn from ‘Agile at Scale’ this months HBR (May/Jun 18)

This month Harvard Business Review (May/Jun 28) contains some great articles.

Managers can’t be great coaches by themselves

  • Coaching is an essential part of employees personal and professional development.  Interestingly, the authors found little correlation between time spent coaching and employee performance
  • The researchers found four types of a manager (see below), with ‘connectors’ being the most effective.
    • Always on managers provide a constant stream of advice.  Working with these managers can be overwhelming and detrimental.  These managers focus on areas they prefer, not what the employee needs and they don’t know their own limitations
    • Teacher managers coach from their personal in-depth experience.  While they provide excellent technical or tactical advice, they struggle to help the employee grow in areas outside their expertise
    • Cheerleader managers offer hand-off support and encouragement but are less proactive and shy away from open and open feedback
    • Connectors managers are the most impactful. They provide tailored feedback and help employees make a connection to someone that can help them.  Connectors know where they have the expertise and where they don’t, and also
  • Focus less on frequency and more on depth and quality.

I found this article pretty compelling.  What shocked me was that my ‘always on’ approach is not the best and I will now work harder to coach within my expertise and connect my team with the right people to help my team grow in areas outside my expertise.

Strategy for start-ups

A go-to-market strategy for any innovation includes four decisions:

  • Which customers to target
  • What technologies to apply
  • What organisational identity to assume
  • How to position the company against competitors

If you already have a company or product, you can support answer these questions with your existing data.  Start-ups are unlikely to have access to the right data.  Therefore, this article argues that start-ups should analyse four possible strategies, to help them formulate the right approach.

The authors argue that you need to make two significant decisions:

  1. Collaborate or compete.  Do you work with existing players or set out on your own?  Collaboration provides access to markets and supply chains, which will allow you to enter a more significant, and the more established market quickly, but there are risks.  You will have less power than incumbents, and you are likely to be delayed by their bureaucracy.  Competition means that you are more able to design and build from scratch and focus on underserved customers. However, established players have more resources, and if they see you as a threat, they can quickly mobilise resources to squash you.
  2. Build a moat or storm a hill.  Tesla shared its IP (early battery patients) to jump-start the electric vehicle industry.  By creating an industry, it created a bigger potential market, reduced the ability of incumbents to squash competitors and reduced the negative impact of network effects (by creating more charging points) – a great example of storming a hill.   The alternative would have been for Tesla to focus on protecting all of its IP, the market would have grown more slowly, but Tesla would have been able to capture more value.

The entrepreneurial strategy compass below (p48) helps crystallize the two choices and the four options.


 The authors suggest that you analyse the four decisions for each quadrant.  This analysis should show which, if any, options are viable.   If you have multiple options, the authors suggest you choose the option that best aligns with the vision and nature of the start-up.

This article provides a compelling framework for analysing potential strategic options. While I believe the analysis is worth the effort, I worry about the suggestion that start-up strategy is deterministic (do X and be successful).  I feel what is missing from this analysis is the recognition of the iterative and exploratory nature of product market fit.

It’s not about the framework

The central thesis if this article is that you cannot apply uniform logic to the problem of creating a business.   The authors argue that post hoc analyses of successful startups are usually very biased, do to hindsight and a lack of evidence.

So what is their advice?

  • Non-plan plans do not work.  Frameworks, such as the entrepreneurial strategy compass (see above) and business model canvas, fail to provide
  • Learn by doing.  Rather than rely on historical data (which helps show why something was successful) create new data.  Create a new product and just get it out there

This article provides a great counterpoint to earlier  ‘Strategy for startups’ article (above).   While separately they provide alternative views, together these two articles offer insightful advice to entrepreneurs.

The surprising power of asking questions

Lawyers, journalists, and doctors are all trained to ask great questions.  Why are the rest of us are not taught to ask better questions?

The authors advise:

  • Ask questions other people will enjoy answering
  • Your approach should depend on whether you are having a competitive conversation or a cooperative conversation
  • In a competitive conversation, your partner may be reluctant to share information and may lie because if they answer your questions, it could put them at a strategic disadvantage
    • Tactics for asking questions in a competitive conversations include:
      • Ask direct yes/no questions
      • Ask detailed follow-up questions
      • Frame questions pessimistically (“X is not going well, right?”)
      • Ask the most sensitive question first
    • Tactics for answering questions in a competitive conversation include:
      • Prepare.  Think in advance about the information you do not want to share
      • Answer a similar question instead of the actual one asked
      • Pose questions in return to gain control of the conversation
      • Share negative information (strategically) to build trust
  • In a cooperative conversation, your partner may shy away from conflict or not share bad news, answering these questions too honestly,
    • Tactics for asking questions in a collaborative conversations include:
      • Ask open-ended questions
      • being with the least sensitive question
      • frame tough questions negatively
    • Tactics for answering questions in a collaborative conversation include:
      • Avoid being boring, use energy, humour and storytelling
      • Don’t dominate the conversation
      • deflect tough questions with your questions or a joke
  • The article also provides a list of good questions:
    • What am I not asking that I should?

The main point I take away from this article is to explicitly decide on what type of conversation I want to have (competitive or collaborative) and adjust my approach.  I also loved the ‘what am I not asking that I should’ question.


Structure that’s not stifling

In this article, Ranjay Gulati attempts to show how you can provide essential direction without shutting them down.

There is an inherent tension between employee empowerment and operational discipline.  While much management literature has attempted to reconcile this with matrix structures or self-managed teams, a clear answer has yet to able to be reproduced.

The author’s central thesis is that guidelines are not the death of freedom if they’re well designed and well implemented.

  • Freedom requires some form of constraints, ‘freedom within a framework.’
  • Freedom means trusting employees to think and act independently on behalf of the organisation
  • People want choice (reinforced by Daniel Pink’s work on motivation) and to be able to express themselves
  • Freedom means responsibility –
  • Freedom means having the right information, understanding and context
  • Freedom needs support from the top.  You have to prevent empire building by functions
  • The degree of freedom is dictated by the industry, but those that are at the leading edge of freedom in their industry are more likely to be high performers, e.g., Revolute, Netflix, Alaska Airlines (probably because they are able to attract and retain the best talent)
  • To define a company-specific framework, you need to work with the whole organisation to establish:
    • Cleary define the purpose.  A single shared goal that sums up the ‘why’ or the organisation.  This needs to give meaning and direction to everything the company does
    • Clearly define priorities.  Behavioural rules that reflect the organisations’ goals.  Simply put, how do employees make tradeoffs, e.g., speed or utilisation
    • Clearly define principles.  Which apply most of the time
  • Implementing the framework.  If the framework was co-created by the whole company, then you can trust the organisation to implement it, but it will be supported by the executive team.
  • The primacy of process is a bad thing.  Processes get refined, improved, more detailed and while this can help optimise routine work, the more specific and detailed the less able to adapt.  Strong process means that people stop thinking, questioning and improving, and just follow the process (see Jeff Bozos letter to shareholders here)

This is a fascinating article.  Accelerate has shown that with digital teams you should er more on the side of empowerment and letter people and teams use their tools of choice. 


Marketing in the Age of Alexa

In the not too distant future, the majority of our daily interactions will be through AIs.  The rise of AI interfaces is likely to disrupt the existing in branch, web and mobile channels.

Niraj Dawar and Neil Bendle suggest that companies will have to shift from marketing to customers to work out how to influence the platforms and AIs.

Platforms will focus on building trust by focusing on two key areas:

  • Accuracy.   Platforms can only convince people to change behaviour (e.g., adopt a new brand) if the platform has consistently reccomended changes that have had a positive impact on the customer
  • Alignment.  Customers need to know that the platform is aligned with their best interests.  If the platform just provides products and contents for the platform to make money, then customers will quickly go elsewhere

Once you understand the motivations of the platform, then you can ask yourself three fundamental questions:

  1. Who is the platform working for?  Successful platforms will create value for
  2. What do we want from the platform?  Look beyond price, platforms can provide data on customer behaviour and price.
  3. How can we be sure the platform chooses up?  We can pay the platform (push) or be in high demand by customers (pull), the second is more powerful.



I feel that this article is ahead of its time.   While many legacy companies are still making the transition to digital and social marketing, innovators are already looking to voice and AI platforms.  Understanding what platforms are trying to achieve (accuracy and alignment) will enable you to create a compelling offering for both the platform and the end customer.


Agile at Scale

New companies are being born agile which gives them a considerable advantage over established players.  Few legacy companies have adapted well, those that have been successful have used Agile to become Agile.

The Agile approach to becoming Agile:

  • Create a list of all of your potential opportunities
  • Create a clear and transparent prioritisation
  • Split priorities down to different teams
  • Set clear goal and success metric, fund these priorities like a VC
  • Leadership teams need to instil agile values through the enterprise, including the parts that do not organise into agile teams

While the original Agile manifesto and principles give little guidance on the structure, size, composition or approach of teams, some consistent themes that have emerged from successful organisations:

  • Self-governing rather than chain-of-command
  • Teams and their leaders need to:
    • Have a major business opportunity
    • Be responsible for a specific outcome
    • Be trusted to work autonomously
      • Clear decision rights
      • Properly resourced
    • Be committed to agile values, principles, and practices
    • Are empowered to work directly with customers
    • Know how to create prototypes, MVPs, and fast feedback loops
    • Be supported by senior executives, who remove impediments, support the team and help the wider organisation adapt

There are many formal frameworks for scaling agile (SAFe, Scaled Agile, Less, DA).  While these frameworks are great for inspiration, they tend to focus on the process and can be very dogmatic.  This article does a great job of providing an excellent, open, high-level framework.