This blog post focuses on how to gain value and competitive advantage from the quality data you already collect today. But for this to work it’s essential that your quality data is stored digitally in a database. If you are not already collecting data digitally, take a look at this blog post that looks into the benefits of digital documentation over the traditional paper form approach.
Is my quality data digital or dead?
If your quality data or as-built documentation looks like the picture below, then I’m afraid your data is dead.
In addition to handwritten paperwork, I would even consider PDFs, text files, spreadsheets, and emails to be dead data too. What separates dead data from digital data, is that it cannot easily be queried or analyzed at scale. This is, of course, regrettable. Especially because your business is into Industry 4.0, Artificial Intelligence and IoT… Right?
OK, there is a way to turn dead data into digital data, but it is a longshot that requires investing serious time and money. Optical character recognition (OCR) services such as Microsoft Azure’s cognitive services or IBM’s Watson AI services can convert paper documents into digital data. Over recent years both output quality and costs have improved a lot. However, I seriously doubt that most businesses (especially SMEs) have the skills and budget to make a full-scale paper-to-digital conversion.
What’s the path forward then?
I think it is probably better to leave the past behind and focus on your future quality data. Two of my previous blog posts provided actionable input on how to design digital quality forms and how digital data collection can increase productivity. The posts provide some inspiration and perhaps the wake-up call you need to turn all your documentation digital.
If you finally decide to move away from paper and over to bits and bytes, take a look at how Prescience can help you digitalize quality assurance, as-built documentation and perhaps even inspire you to start measuring quality performance systematically.
Assuming that you make it. Assuming that you manage to start capturing quality data digitally going forward, here are some key ways to exploit the quality data you collect to gain value and competitive advantage for your business.
1. Improve traceability and customer loyalty
Traceability probably sounds obvious to you. Actually, traceability is most likely why you started collecting the data in the first place, right?
Industrial businesses spend tons of resources documenting raw material, procedures, parameters, measures, operator certificates, etc. to provide traceability. Just in case something goes wrong (God forbid) and we need to hold someone accountable. This is why customers, and sometimes authorities, require documentation.
Traceability is excellent, but for me, traceability isn’t just traceability. It actually matters how well businesses can execute traceability. Firstly, in terms of how businesses capture and manage data. Subsequently, how they process data when something does go wrong. During my years in the wind industry, I have seen both good and bad examples.
Poor traceability can be an order disqualifier
Strong traceability processes are often an order qualifier and a mandatory step in the supplier approval audit. Oppositely, poor traceability if an order disqualifier and most definitely a long term order looser.
I would argue that most paper-based traceability gives a false sense of security. Mainly because dead data and poor data management weakens traceability. I’ve experienced simple traceability inquiries that clearly demonstrated a lack of control. More importantly, the subsequent containment effort to locate additional cases was a complete disaster, providing very-low-confidence results, to say the least. This was due to poor data management and the inability to cross-reference data. After this experience, our sourcing team was reluctant to ever use the supplier again.
Luckily, I have also seen suppliers who dominated traceability. They were able to quickly access the data source and were able to (with a little bit of work) query the data and correlate it with additional meta-data. This traceability analysis projected confidence because it provided an accurate result. Moreover, it also labeled the supplier as a trustworthy partner who took traceability seriously.
The lesson here is that businesses who dominate quality and traceability gets repeat business. Dominating quality and traceability requires digital data collection, storage, and subsequent traceability analysis. A basement full of cardboard boxes with manual paper forms cannot provide fast, accurate and credible results. The risk of overlooking something is simply too great.
2. Quality data can improve the customer experience
Our users highlight customer demands for more and faster quality documentation as a major challenge. However, digital access to quality data enables suppliers to tailor improved customer experiences where customers can serve themselves at their own convenience.
Prescience is an excellent example of this. The Prescience dashboard displays new quality data in real-time as inspectors provide it along with the overall data collection progress. This basically allows customers to take a look for themselves, instead of sending emails to ask for documentation.
Similarly, customers can access quality data via the Prescience Application Programming Interface (API). This would allow customers to electronically export quality data. It would even enable customers to develop their own self-service solution to display quality data based on e.g. a serial number.
This will dramatically reduce the internal resources and costs required to answer customer inquiries. More importantly, it will provide what customers are really asking for; CONTROL and TRANSPARENCY. Surely customers will look favorable upon suppliers who can provide such an improved buying experience.
3. Quality data can drive improved performance
It is an old truth that you cannot improve what you don’t measure. To expand on that cliché I would say that most businesses who do measure are not really improving. Logging the data is simply not enough. There must also be a systematic and effortless method to turn data into compelling charts and KPIs. And this is where databases beats paper and spreadsheets each day of the week and twice on Sundays.
The capability to systematically track performance is a game-changer for the quality team. Being able to zoom in on all the data for just a single data point can provide a lot of insights:
- How are shifts, production lines or factories benchmarking against each other?
- How is performance developing over time?
- Which improvement efforts yield results? (and which don’t?)
- Can we attribute poor performance to specific people, factories or item numbers?
Digitalizing quality data provides many opportunities to strengthen performance measurement, without doing a ton of extra work. If you believe your supply chain could benefit from performance measurement, take a look at how Prescience can help implement quality performance measurement.
4. Quality data can provide new insights
Great, so we established that the quality data can help the quality team to improve quality performance. But can it also help other stakeholders? Absolutely! Allowing other stakeholders to combine quality data with their own business data can yield significant results.
My favorite real-life case was a cost-out project to reduce paint overconsumption. At the beginning of the project, we asked suppliers to log paint buckets as a way to measure consumption. However, this approach did not work well; a) because paint buckets overlapped components, b) operators constantly forgot to register, and c) it was simply too difficult to measure consumption accurately on a global scale.
This was when someone got the idea to use the paint thickness data instead. Suppliers measure this systematically as proof of the paint manufacturer’s corrosion protection warranty. By correlating this data with the paint manufacturer’s product data sheet we could approximate paint consumption.
This suddenly provided a dataset covering 100% of components going several years back. It also meant that data collection going forward would require zero effort from us. Using the data we could easily identify underperforming suppliers and track whether they improved over time. Similarly, it allowed us to attribute overconsumption to specific equipment and operators.
Combined with manufacturing data we were able to benchmark suppliers and track performance evolution over time. We were even able to attribute poor performance to specific painting booths and painters. As a result, improvement efforts could be targeted at poor performers with pinpoint accuracy.
In conclusion, all this quality data can serve as valuable input to someone else. If you can attribute poor performance to specific suppliers the sourcing team needs to know. In fact, such metrics should be part of the strategic supplier selection process. Similarly, if you can attribute poor performance to specific item numbers the design team needs to know. This is really the input they need to develop the improved or entirely new products that will drive revenue in the future.
5. BI, ML, and AI is a non-starter without data
BI, ML, AI, Huh?
Yup, digital quality data is a perfect match for Business Intelligence (BI), Machine Learning (ML) and Artifical Intelligence (AI). In fact not having digital data is what prevents most businesses from using these new technologies for quality purposes.
Many people still think BI requires the IT department to be on board. This is actually no longer true. Self-service BI solutions such as Power Bi, Qlik, and Tableau makes it incredibly simple for regular employees to connect to datasets and make explorative analysis.
Being able to make such drag and drop analysis on huge quality datasets have never been easier. This is a fast and flexible way to confirm a hypothesis or get new insights to drive efforts and decision making. At the end of the day, this is really what I think defines a data-driven business. Being able to quickly model and visualize data in news ways that we previously did not imagine.
Machine Learning and AI
This is an area where I have less (almost no) practical experience, but where I have high hopes and expectations. The emergence of services like Google TensorFlow, Microsofts Azure Machine Learning and IBM Watson has really reduced both the cost and knowledge entry barriers.
The main obstacle here is really the lack of high-quality datasets. Machine Learning and AI both use past experiences to train and get smarter. But without a historical dataset on which to train it is really difficult to achieve results. Having a large digital quality dataset is a MUST before businesses can venture into these technologies.
If you are one of those businesses that keep quality data in cardboard boxes in the basement, you are missing out on some important opportunities.
It is clear that digital access to quality data indirectly drives competitive advantage. Being able to query, model, convert, analyze, export and share quality data enables businesses to:
- Offer superior traceability that inspires customer loyalty
- Deliver an improved buying experience customers will love
- Track quality performance systematically at almost no effort
- Gain insights and metrics to support improvement in all areas of the business
- Finally, digital datasets enable businesses to tap into BI, ML, and AI.
If you aren’t already running digital or even planning to do so, I hope this blog post has been an eye-opener. There are many ways digital quality data can help your business improve. At least you should examine the cost/benefit of going digital, instead of just discarding the notion.
I am sure you have more ideas or practical experiences for utilizing digital quality data to drive business improvement and competitive advantage. I would really appreciate your feedback in the comments 🙂
Have a good one!