So it's interesting to me to read a completely reasonable list of "things to do" before implementing AI or any of the new digital transformation toolkit and think: that's absolutely correct in theory and likely 100% wrong. I thought that when I read the article entitled 3 Steps to Gear up for AI and the future of work. The advice is meaningful and probably useful, conveyed thousands of times about new technologies or new approaches. It advocates:
- Determine a Use Case for the new technology or approach
- Train people to be more proficient users of the new technology
- Start small - find small successes
- ERP
- Lean
- Agile
- Six Sigma
- Doing business on the internet
For innovation and digital transformation, perhaps we should be a bit more unreasonable.
Digital Transformation
First, we've skipped a step. We need to define what digital transformation is, or what we think it is. I'll define it as the implementation of a number of technologies (like AI, machine learning, blockchain, IoT, robotics, big data and so on) which transforms business processes and strategies. Since each technology has a multitude of potential use cases, from generating new products and services to increasing revenue to cutting costs, we should be more circumspect about what the value proposition and use case is. And, since most new technologies enter the mainstream market by creating greater efficiencies, perhaps we should more accurately ask: what significant customer need can it address, or what significant inefficiency or cost can it remove? Further, how important or significant should the need or opportunity be?
They don't call it digital incrementalization. They call it digital transformation. The same is true for innovation. We need to be thinking bigger. The time to conduct small experiments around the fringe is over. We need to use approaches and tools like digital transformation and innovation to solve important and urgent efficiency and cost needs, or to resolve major customer needs. There's your "use case".
And yes, I've combined my response to the "use case" point and the "start small" point into one lengthy paragraph. If by now your company cannot use innovation or digital transformation to do some big things, to create interesting new products or business models, or to radically transform the customer experience, the end is near. The time for starting small was 5-10 years ago while innovation and digital transformation were still relatively new. You cannot compete with other firms that are doing much larger or bolder experiments because the cycles of learning and implementation are collapsing.
John Boyd said it best in his OODA loop (Orient-Observe-Decide-Act). If a competitor can progress through this loop faster than you can, you are a target, not an adversary. Starting small, experimenting and then scaling up is valuable when the technologies or capabilities are still new. When everyone is doing them, it's time to make bigger bets. Corporations need to be more agile, yes, and experiment, yes, but also must be able to scale quickly and implement quickly.
Training
The article also advocates training. There are several concerns I have with training, on new technologies such as AI or blockchain, or on innovation tools and methods. While it always sounds appropriate to invest in training your people on tools and methods you need them to use, too often training isn't valuable or worse, it is wasted. There are a couple of reasons for this.
First, training is the first corporate expense to get cut in lean times. While deep, formal training is often valuable, it is also often hard to schedule and hard to justify. Training is never "top of mind" for people who are implementing technology, and is often difficult to acquire and schedule for those who should be using the new technology.
Second, training is often wasted in innovation work because we train people on innovation tools or methods but send them back to do their regular jobs. If you want innovation training to be valuable, you need to train people on tools and methods they use immediately after the training. Otherwise, don't bother, because the regular work cadence will soon cause them to ignore or forget anything new they've learned if they don't use it.
Third, training for digital transformation tools is also a bit difficult, because you first have to identify which method or tool you'll implement, what problem or need you are trying to solve, and how the method or tool creates insights or data or simplifies a process to solve the problem. In many cases digital transformation may simply remove people from an activity, so the training they may need is training in a completely new role. Other training requirements include the ability to implement the new technology - but it's probably better to hire this work rather than train people for it, and to be able to understand and assess the information that the digital tools create. It's probably this last idea that is most useful, but also most contextual. The insights that each digital technology provides and how the data is interpreted will be different in case by case basis. It will require people with excellent data interpretation and contextual skills to interpret well.
What instead?
What if the right answers for innovation and digital transformation - at least at this point in history - aren't start small, train users and identify use cases - but instead the new conventional wisdom should be:
- Address wicked problems head on with the goal of completely solving them, focusing either on radical cost reduction or dramatic improvements in customer experience
- Think differently about training, in 3 dimensions: buy the experience you need rather than train for it (technology), train people to understand the information presented by new tools and methods, prepare to train people for new jobs and roles once their existing jobs are eliminated
- Start big and go bigger - the pace of change doesn't allow small, continuous experiments when technologies and capabilities are relatively mature. Go big or stay home.
0 comments: