Overcoming digital difficulty #7: Neither top-down nor bottom-up works
Blog: The Digital Enterprise
“Most of what we call management consists of preventing people getting their work done” (Peter Drucker). In digital transformation, management need to ‘get out of the way’ and create space for innovation, yet at the same time provide sufficient direction to ensure that the innovation that flows will deliver the desired outcome. It’s a tough balancing act. This is the 7th post in a series building on 10 reasons why digital transformations are difficult. Here are some proven approaches to balancing top-down and bottom-up:
- Adopt new models for driving change. Traditional projects are run like an infantry division – highly organised but hierarchical and slow. Effective digital project teams act like a special forces operation – fast, autonomous and agile.
- Fund like a VC. Successful digital transformations often use a VC-style funding model, investing in a broad portfolio of projects. Many projects receive seed funding, but only those that show promise are allocated further funds to build, then deploy a product.
- Flatten the organisation. Reducing the number of management layers between CEO and front-line leads to better communication and faster decision making. Some firms (e.g., Zappos) are pushing the boundaries by avoiding formal hierarchy altogether. That’s a step too far for most organisations….but there is plenty of scope to flatten.
- Strengthen the architecture function. The key risk of having multiple independent projects is that end products (e.g. user-experiences, processes, data models) don’t connect to each other. A strong architecture function that spans business and IT can ensure that all the pieces together make a coherent picture. Crucially the Chief Architect needs the right tools to exert control, for example a veto over every investment project.
In summary then, top-down vs. bottom-up is an increasingly redundant paradigm. New organisation, delivery and funding models are needed to bridge the gap.