4 keys to successful implementation

21 March 2019

Your business will only benefit from data science and data-informed decision making after successful implementation. Before embarking on the implementation journey, think about the following:

1. Balance

The decisions made can only give you an advantage if you have as complete a picture as the data can provide. Consider what data sources could be used; internal as well as external (it is becoming easier and easier to source external data – most governments, universities and even major charitable organisations publish a remarkable array of data sets). The data should give high-level insights and the ability to drill down to the detail to make operational changes that won’t have unintended consequences (this is where scenario modelling can be helpful). You also need a balance of old and new data; as you implement changes your customer and client behaviour might change so new data will need to continue to feed the models.

2. Openness

Insights can be unexpected and even be confronting in some instances. Is your business ready to be challenged by these insights? The challenges can be to pre-conceived business ideas or ways of doing things (machine learning systems could potentially automate certain operational decisions and actions). Unless your business culture is open to acting on new insights, you won’t benefit from the implementation. Don’t give up – over time you might be able to move people to accept the new insights and act on them. I have been involved in organisations where insights were talked about for two to three months before they were accepted. It was frustrating to see the opportunity to take action slip away, but in my experience it is almost an inevitable process that needs to be worked through in any organisation. Be prepared that taking action based on insights might take longer than you thought it would.

3. Integration

This is where the true power lies. With advances in technology, more advanced analysis based on a variety of data sources is possible. It might still take a while and quite a bit of effort to clean and prepare unstructured data, but it is worth it to get richer insights.

4. Honesty

Although you need an open culture, you also need to be honest about the limitation of data science. Remember the insights are only as good as the data available . You need to know where the data quality issues are, where the gaps are and also think through the business implications.

Data science and data-informed decision making is such an exciting area and there are so many potential applications. Choose areas in your business that are ready and will most benefit from implementation to be part of the pilot.

Good luck with the implementation!

 Further reading:

Artificial Intelligence – What it is and why it matters

Artificial intelligence unlocks the true power of analytics

The incredible inventions of intuitive AI (TED – Maurice Conti)

Barriers to AI adoption – culture, fear, lack of talent, not integrated with strategy


Close Menu