CIO - move from collecting data to utilising it

With regard to artificial intelligence and machine learning, there is a common illusion a single technology, software, or algorithm would help a company boost its business. As if there would be some sort of a black box, which digests all the data and spits out an operational guideline.

Sorry to disappoint, there is no black box. There has been no new mystical breakthrough in algorithm development, and no software can standardise data processing and predicting the future. Unfortunately, results that really benefit the business are not created by connecting data sources into a crystal ball.

However, getting started is not that complicated. The revolution in using data is based on increased computing capacity and open source development. Companies can set up a powerful computing environment in minutes, and with reasonable costs, if they are willing to use cloud services. The algorithms needed for data processing and predictive analytics are already freely available.

The good news, then, is that the that business managers don’t have to whine about various technology choices. You don’t even need expensive licenses. All you need is an expert, who knows how to use and modify algorithms to suit your business needs.

Rule no. 1: Be impatient

In order to reap benefits from analytics, you do not need large projects spanning many years. Although it’s tempting to start by planning a big project with all the data neatly available to the entire organisation in a structured form, it is more fruitful to take a different approach. How can we quickly benefit from our data, and how can we use it to accelerate our business? Modern technology is modular and, when properly designed, also flexible. It is easier to plan an organization-wide project, once there is an understanding of the benefits to be achieved.

Rule no. 2: Tolerate disorganised data

One often hears we cannot do this until the data is in better order. We need precise specifications. Our data contains errors. These must be fixed first. It is, of course, useful to get the company’s data sets organised. However, this should not slow down development, as an experienced professional will be able to correct erroneous data in no time. A proficient expert can write algorithms, which for example reconcile different data entry practices and estimate missing data based on some other characteristics.

Rule no. 3: Simplify and succeed

Business is fundamentally about predicting the future and taking action accordingly. The point of artificial intelligence and machine learning is not in the use of brilliant algorithms, but rather in the ability to simplify key business questions into mathematical models. The winner is the one who can formulate the right abstraction for the key question – in a way that it can be answered.

Rule no. 4: Acquire expertise that will benefit you immediately

Almost without exception, the best option is to quickly implement a Proof-of-Concept or a smaller solution, which yields immediately measurable business benefits. This is, of course, easier said than done! However, an experienced professional is, together with business experts, able to find a suitable starting point. A consultant is not the best to advise on how to do business. But she knows the possibilities of using data and is able to harness large amounts of data in such a way it is useful in decision making and running business.

Conclusions

When it comes to buying analytics, the choice is clear. You should choose a competent data scientist who understands the customer. At Nitor, we focus on resolving the customer’s problems in a business-oriented way, with tools best suited to the task at hand instead of relying on technology-driven solutions. If you would like to know more, we’d be happy to come and start charting potential development targets and creating business value for you.