Strategists can be obsessed with spotting new opportunities. Yet this is of little use if the organization is unable to seize them. Most strategies fail because they are not backed up with the right organizational competences, detached from the operational reality, or simply not understood by staff. Your biggest strategy problem is typically a disconnect between the upper echelon and the rest. AI can help to overcome this.
The public debate about AI has focused on whether machines will replace humans. This is hardly surprising after the launch of large language model like ChatGPT. But what if the main effect of AI – at least for now - is not replacing humans, but creating more efficient connections between them?
For strategists this is an attractive proposition. ChatGPT and other large language models do not have the discretionary power necessary to craft creative strategies. But the possibility of instantly processing millions of data points could be used internally. Executives can use this to listen to all their employees and figure out which strategies are achievable.
There is an added bonus here: talking about strategy increases the likelihood that staff actually know the plan. A recent survey showed that only 18 per cent of middle managers can correctly name their company’s top three priorities. For frontline supervisors the number drops to 13 per cent.
AI can be applied in a potentially game-changing manner using three different techniques: in-house social media, strategy jams, and strategy dialogues.
In-house social media
Social media has long been mainstream in corporate America and a growing number of companies now use it to engage staff around strategic matters.
Unilever uses it in the build-up to its annual leadership summit, while Telefonica executives realized its power when staff started to chat about customer engagement. Giving employees a space to talk about strategy has an obvious benefit, as they can figure out what it means for them.
However, it has been harder to figure out which ideas are widely viewed as promising and which issues prevent a successful strategy execution. Something mentioned on a chat might catch the ear of the CEO but this is not the same as a systematic analysis.
Topic modelling and other machine learning techniques can provide a robust digestion of large datasets, making the cognitive biases of executives less of an issue.
Strategy jams
A strategy jam is akin to a large online conference where thousands can debate strategic issues in different streams. IBM pioneered the technique using its Watson computer to facilitate both the conversations and analysis.
Three things are important here. First, the limited time frame – typically three days – concentrates minds. It’s an event!
Second, human moderators are crucial to maintain engagement and make sure that everyone stays on topic. Much of the success is hidden, in the sense that people in different pockets of an organization can figure out what a new strategy means for them and connect with other departments they don’t usually talk to. The moderator facilitates this, but also uses AI to stay on top of things.
Third, AI tools tease out the essence of the jam once it is done. A “Debator tool”, for example, identifies the main arguments for and against a particular issue that became prominent during the jam. Strategy execution can then be aligned with the sentiments of the organization. Knowing exactly what people think, rather than guessing, works.
Strategy dialogues
While social media enables ongoing interaction and jams limit it to a few days, strategy dialogues are set up in two phases. Phase one starts with an open question (or several). Something like, “We want to grow our retail business 10 per cent this year. How do you think we can do this?” Or, “The Board recently approved our strategic priorities ABC. What do you need to execute in your area of influence?”
CircleLytics, the Dutch firm who developed this tool, noted that 95 per cent of the answers fall between 150 and 190 characters and decided to set an upper limit of 220 characters. Enough to say something substantive, but not long enough to overwhelm anybody.
Three days after the answers were collected, participants log back in. The founder Maurik Dippel notes that “this is essential to maximize people's learning from new perspectives, take time to reflect, and listen this way to many others”.
In phase two, each participant sees 15 answers from other people. AI has picked responses which stand in contrast to the participant’s own answer and also present substantial variation. The underlying idea is that we increase our horizon when we are confronted with suggestions that are not aligned with our own thinking. Participants then rate each of these 15 answers from -3 to +3. In addition, they can highlight words that seem important and explain their reaction.
Once all the data is collected AI is used again, identifying the top and bottom five answers. Executives then know what staff collectively think.
AI supported engagement does not mean corporate democracy
The power of AI can help executives to better listen to the organization. They might be worried that this takes decision making power out of their hands. But as long as they are explicit about what staff can expect, this is not turning into corporate democracy. It simply helps executives to set execution methods that are in-tune with what’s possible.
"strategy" - Google News
June 13, 2023 at 04:42PM
https://ift.tt/YnJ2V7a
How AI Can Solve Your Biggest Strategy Problem - Forbes
"strategy" - Google News
https://ift.tt/5xtTezo
https://ift.tt/SBcpNoU
Bagikan Berita Ini
0 Response to "How AI Can Solve Your Biggest Strategy Problem - Forbes"
Post a Comment