PLEADING FOR MORE FAIRNESS IN DATA COLLECTION: WHY ORGANIZATIONS OFTEN GET THE DATA THEY DESERVE – NOT THE DATA THEY NEED

Why is there still so much ‘bad’ – not usable – data out there in companies? One reason: the generation of data is still a weak spot in many organizations. This goes back to a lack of understanding for the motives and drivers of the people that are asked to submit their data, Wiebke Apitzsch, Managing Director of consulting boutique TTE Strategy argues. Her approach: Data needs to be shared back to the submitter as soon as possible with added value. Not only does this keep up motivation. It ensures high data quality – as only those who know the reality on site can tell if the data reflects it accurately.

How do you feel if someone asks you to give him something, and you do so, and you never see it again? And then he keeps coming back, asking for more??

Well, unless you are an extremely generous and giving person, you might get irritated, right?

And while there is a common understanding that the above is unacceptable behavior in private life, it has become a common process in business when it comes to data requests.

OK, you may say, people get paid for work, so how can you even discuss the need of payback?

There are two very different kind of tasks at work. The core, and the extra. To explain this, let’s do a little case. You are the boss. It is end of year and you review the team.

• Engineer A fixes every broken machine in no time and even comes to work at night to help his co-workers if they get stuck. But when it comes to documentation, he does not use new possibilities: He just does not switch from paper documentation to the new tablet.
• Engineer B struggles in front of a machine, even after a few months in role. At the same time, he fills each new electronic form with utmost care and all the detail.

What do you do?

Now, likely, you keep A, and discuss opportunities for B to show his skills in a different setting.

A performs in the core, B in extras. But core remains core, they are not equal.

That is: asking people to generate more and more data, means, that you ask them for not-compensated, extra effort.

You may argue that the data requests are there for the overall good of the company, and hence the team should have an intrinsic motivation to do it. But then remember that COVID, not the threat of climate change reduced air travel worldwide. Only very few people will constantly invest bigger goals if they do not see an immediate effect for themselves.

That said: How can we make sure that even Engineer A decides to fill the new forms, even though it means additional effort for him, he is not fired if he denies, and he personally dislikes computer-work?

The answer is pretty simple: Engineer A will fill the form if it generates direct value for him.

How to create value for those who share the data?

The first thing you should crosscheck in your transformation strategy is: how does the operating team benefit from the new transparency?

You can ask them, but it is unlikely that you get a clear answer out of the blue. A better approach it, to analyze opportunities together. Analyze how does Engineer A work. And what knowledge could help him to save more time than what he invests for the input. Good questions can be:

• How does he plan his work each day?
• What information does he collect before executing his tasks?
• What does he wish he had known before he started to work on it?
• What are topics he and his peers discuss when they hand over from one shift to the other?

Likely, there is a lot that technology can do for him. Like having a standard visualization for machine health. All SOPs at hand, including the specialties of this installation. The contact details of all relevant supplier-helpdesks just one click away. If you manage to build up such clear and relevant reports with the team, that they enjoy using them, then your project will be a success. If the shopfloor team uses the data, peer pressure will force the last colleague who is still reluctant to change into the process.

Quality in, quality out

Diligence itself is not the only reason why dashboards for the team should be build first. You could, if needed, get this done with pressure as well. At least to some extent.

But there is a second effect, that is equally important: Data quality.

Be it manually entered or machine-generated data, not in one case in my career I have ever received a dataset that was ready to use. Leave alone if the data was generated in different locations. There are many reasons for that, of course. But one is, that if you do not see how the data is used, it is a little bit like with these old data input forms where you could write as much as you wanted to, but you only see the last 20 digits. The story get’s weird. And next time, with no historic data available, chances are the information loses consistency.

In contrast, if our operators have a dashboard in front of them, showing only 50 percent of the normal spoilage, but seeing, the bins are full as usual, they instantly know that something is going wrong. And, as this messes up their new handover procedure and their boss will soon show up to ask what happened, they have a very good reason to highlight the issue and fix it.

Data quality starts with good data, but it ends with people validating them. Most organizations need to decide crucial strategic topics on a data-driven information base, which is the reason for the high importance of a proper data quality. And why invest too much time into data cleansing afterwards if you can avoid getting messy data in the first place?

Sharing back the data with those who know it best makes sure, that it remains valid.

And if the data is not accurate, executives find themselves in never ending discussions. Some mistakes are presented and proven, now the entire analysis is no longer trusted. Maybe it is not pointless, but much weaker than its potential.

In essence:

Sharing data back with those who generate it as a priority is not only fair. It also generates a constant flow of valuable information and ensures high data quality. It builds up a solid foundation, that can then be used to build up advanced functionality. Only if the groundwork was done properly, advanced Analytics can show their full potential.