Many companies are cold starting their data projects

As mentioned in the previous blogpost, Data Professionals are looking at the door when it comes to their current roles due to a few factors, a major one being companies being unprepared and the new hire having to start things from scratch. This situation is often referred to as ‘Cold Starting’ but why is this happening?

When starting to build a team for projects within data, the first line of priorities are collection including user generated content, external data, instrumentation and logging. Useful, useable data is key for taking advantage of AI. Starting with an assessment of whether the data being collected is the correct type of data to satisfy business needs further down the line. Obviously, this varies and depends on the type of business you run, so whether this is categorical or numerical, this should be in place before you expect there to be any momentum when this data can be assessed.

Some companies fail to understand that these foundational blocks must be in place first and that there can be trial and error before this is all up to speed. What we mean by this is that firstly the requisite types of data must be sought out and then interrogated in order to discover whether the data being pulled through is actually correct. Within payment transactions, for example, there are different types and if the incorrect type is being collated, this can have far-reaching repercussions.

This initial stage, as well as ensuring that the right data is being used, requires resources. This means a leap of faith on the part of the company which, in many cases, does not yet comprehend the true value of what is being proposed. This is what we believe is behind the cold starting. Companies adopting this new approach to their information are unsure, until they are sure, and this results in a rush to get all parts of the machine in place and moving at the same time. Misunderstanding the complexity can result in false starts, a high turnover of staff and the aforementioned failures of data.

From that point onwards, data should be analysed. Analytics means the systematic analysis of stats or data and this can involve many different KPIs depending on a business’s needs. AI should then use algorithms to improve KPis and these improvements will feed back into the system ensuring optimisation and higher profits for the company. Of course, we have simplified it here and there are more moving parts in reality but it is surprising how many people don’t cover the basics when starting.

The company will need someone with the appropriate experience on hand to guide the whole process and ensure that the correct seniority of staff are hired and the staffing levels are correct during all stages of the projects. Another problem that can arise if this does not happen is when a junior data scientist is tasked with creating the system and they use algorithms from practice sessions on real world business issues. Getting the right mix is so important.

Indecision can result from a clash of priorities at C-level between wanting the best and wanting the cheapest solution to the problem. If one or more members are afraid that AI is just the newest fad, which will fade some time soon, then that explains why they hold off on it until they can hold off no longer. This is very true of those industries that are not traditionally data-driven.

Start-ups have an easier time of it because they are often engaged with data during the planning process. Older businesses need to reimagine their business from the bottom up and sometimes that can be a daunting possibility. The truth is that the future won’t wait for a business so the sooner they can get onboard with AI and Data, the better. As the old saying goes, “The best time to start something is years ago. The second best time is now.”

Curious about how Zenshin Talent can help your organisation? Contact us today for a no-strings conversation about your needs and our experience.