Data Insights

My personal journey of data intelligence discovery

Petr Mahdalíček
September 1, 2021 | 23 min read
I have been on a journey of data intelligence discovery these last 17 months during the COVID pandemic, and in this personal blog post I will be sharing my thoughts and experiences of how Simplity made use of this time. The pandemic was initially quite depressing. It was clearly going to affect business for some time, with little gain for many companies including mine and so, I saw an opportunity. After 12 years we are now expanding into new areas of business. Our logic was to invest in our business during COVID and improve what we could offer, but we weren’t quite sure how at first. This is what we did, what we learnt, and why it is important.

I realized that we could adapt our business to help a wider range of industries with the knowledge and expertise we have accrued. For the last 12 years Simplity has worked with very large financial institutions, primarily the banking sector with millions of customers and huge datasets. The valuable experience we have gained, and continue to gain, working with these institutions has taught us that these datasets and business processes are extraordinarily complex. But we can utilize the skills learnt across many other industries. Bigger volumes of data tend to be a technological problem whereas complexity of data can only be unraveled with the help of knowledge and experience.

The pandemic created the need for quality e-commerce strategies and the need for data and systems to be remotely and precisely joined up, all in double-quick time. Our knowledge means this is exactly what we can help companies achieve.

So, as the pandemic grew, we started our research and design to focus on bringing new services to market that could help companies of all sizes and types. These services would improve business decisions and quickly impact return on investment along with boosting revenues. Our in-depth data governance and analytics understanding, and skills, allow us to swiftly assist organizations with their initial data assessments, ongoing data curation, business intelligence reports, customer matching, churn, and loyalty.

And there was more…

Machine learning and data science are of critical interest to all companies now, from small to large. Everyone must work with data. Only those that work with it correctly and efficiently will win out. Data intelligence is essential. Data science without high quality data cannot achieve proper analytics.

The vast amount of knowledge and skills that we have gathered enables us to work with and prepare data ready to combine it with new analytics to help businesses. Working with data is a requirement and we can focus on companies with growing data and the demands that places on them.

We’ve been there, seen it and done it – the troubles, issues, traps, pitfalls and challenges. We overcame them and succeeded. The experience and knowledge we have learnt comes from years of working with very complex data projects for large financial institutions. This is why I believe in what we do.

The next step after data assessment and curation is customer data. This is not just about a database of customers or a mailing list! This is understanding your customers with a holistic 360-degree view. It helps you reduce customer churn and improves customer loyalty. It enables you to sell better and sell more. You can analyze shopping baskets smarter to give you opportunities to upsell and cross-sell, increasing revenue from existing customers. You can improve the customer’s experience too, making it more personalized, increasing their loyalty, and reducing their likelihood of leaving you for a competitor.

During the last 17 months, we have developed new services based on these elements. We have a new, rebranded website with examples of how we proved that what we researched worked with real-world clients and their customers e.g., Mojekolo (a large bike retailer and e-tailer).

My journey of data intelligence discovery during the pandemic to adapt our business to help a wider range of industries, found that the key to customer analytics is data preparation that builds the foundation for the data science that follows.

As always, if you have bad data in the first place, you will get bad results. Young companies with modest experience might not have the full understanding. They may be fast, agile, new, and whizz-bang, but they may not be able to help you prepare, qualify, and manage your data properly. Thus, the results you get will either be bad or simply, not good enough compared to what you could really achieve.

Do you trust your data?

QUESTION: When COVID started impacting the world, which European country had the highest rate of excess mortality during January to June 2020?

The highest "peak" at national level was Spain, but areas of Northern Italy had even higher peaks. But by the end of May "England had seen the highest overall relative excess mortality out of all the European countries compared" – source.

However, for a start, there are three different measures for mortality - age-standardized mortality rates (ASMRs), relative age-standardized mortality rates (rASMRs), and relative cumulative age-standardized mortality rates (rcASMRs). How are the numbers reported for each country, who is reporting it, what about possible differences in measuring the data?

Numbers and data are all about the context i.e., the parameters and the rules. Data quality and stewardship of that data allows companies to define the business rules, control these parameters and consequently trust the data they are using for analytics. It is easy to see that numbers and statistics can be manipulated for the media, but in business, it is a bad way of working because if you can't trust your data, you can't run your business properly.

Datasets can be a puzzle that needs to be solved with sophisticated data management. With big sets of data comes great responsibility – that responsibility may require using the right tools such as Accurity.

I also uncovered the fact that there is a lot of misunderstanding around the need for good data management. Here are some real-world examples:

Do you trust your data?

QUESTION: Do you know all that you need to know about your customer named John Doe? If you answered, “yes, because we have all of his details in our customer database”, then what happens when there is more than one John Doe in your database? In addition, Is the John Doe in your customer database the same John Doe that is in your mailing list database? Or the one in the finance/loans database? Or the one in your customer service/support database? Or the one in the delivery address database? Is John Doe married to Jane Doe that is also in the database? Is John Doe in the customer database the same as Dr. J. Doe in the inquiry database? Master data management like this allows companies to harmonize all the data giving them a single 360-degree view of their customer across all areas of the business and consequently they can trust the data they are using for analytics.

QUESTION: If you’re selling goods e.g., a customizable Power Master 5000 widget, can you track every component in that widget? If you answered, “yes, because the widget is broken down as individual components in our component database”, then what happens when that component has to exist in multiple data locations, not just your component database? Can you track it properly? It will be needed in an online, real-time, stock-dependent sales system allowing for a customized purchase, including downgrade, upgrade, upsell, cross-sell scenarios. But it also needs to be in a third-party supplier’s database that you do not have control over. It also needs to be in your service catalog database so that it can be sent out as a spare/replacement part or used by service personnel/repair stores. Can you track the component across all these systems and more? Reference data management like this allows companies to set business rules, harmonize all the data and consequently trust the data they are using for analytics.

A major problem of data projects is that traditionally it is very easy for them to be long-term, expensive, and ultimately not actually achieve what was originally desired!

Longer projects often fail because they end up being more expensive than expected, momentum is easily lost, the scope of the requirements can change over time, frustration can set in, and thus no one is happy. Shortening projects without losing quality and focusing on specific, smaller goals enables the project to be quicker, more agile, the business can rapidly pivot the desired outcomes, and ultimately it is less expensive.

First, we found that it is possible to break down longer and larger projects into smaller manageable steps, a bit like having a roadmap, but it is also possible to have smaller, faster data projects that deliver real results quickly.

When you work this way there are significant benefits. Initial results can help with calculating return on investment (ROI), narrow the focus of the project, or even allow for re-assessment to pivot the requirements. Taking less time is lower investment and if it still doesn't work out you get to know that fact quicker. And you might discover that it can simply be automated in some way rather than throwing lots more person-hours at it.

For example, a quick project might be to analyze customer churn with a business that wants to focus on subscribers where they are currently losing around 1000 of them a month at a loss of $200,000. The project might not achieve saving the whole 1000 but it could save 25% within 3 months. Statistically, that might not be viewed as a huge success but, and the big but is, that 25% is still 250 customers retained, $50,000 gained, plus valuable data obtained. All of that could be reinvested and pivoted into a different project.

We at Simplity believe in both ways of working but are concentrating on significantly helping businesses with shorter projects as our data intelligence experience and knowledge can help many different organizations including e-commerce and the media and entertainment industries.

We specifically developed services that focus on getting more from existing customers rather than chasing new ones. Why? Because it is significantly cheaper to keep customers happy than to attract new ones to join. Precise personal marketing with a focus on upsells and cross-sales will bring in more from existing customers. It is important to make sure that your customer spends their money with you e.g., in your e-shop, on your services, a continued subscription.

For example, you may have an e-shop and it is doing well because of the pandemic. However, based on Simplity's Customer Analytics using customer segmentation you can discover that according to the items that John Doe bought in your store during the last two months he is 78.3% likely to be a frequent customer. Based on this you can match John Doe against a segmentation profile Frequent Customer and determine that John might also want socks, shorts, wristbands, and shoes (and he could be buying those elsewhere currently). Now you can also start to personalize the offers for all your customers!

During my journey of data intelligence discovery to adapt our business to help a wider range of industries during COVID, I unearthed many elements, but one has become untouchable. Maturity is not always slow, behind-the-times, and boring.

Maturity can actually be a benefit. Some will gather and process data as fast as they can. Some will rapidly populate the data and present it in the swankiest way through reports. But the data that is being used must make sense, otherwise, the quick and cheap data assessment and curation is not actually an investment. The reports that will be produced will look great but cannot be trusted!

Lack of trust in data that is not well and transparently defined, including clear data stewardship, almost always leads to individual data silos within a company being created and maintained. When the data becomes splintered, unaligned, and not synchronized like this, departments stop trusting the data not under their direct control. Parts of the company don't trust data managed and controlled by sales, salespeople don't trust data managed and controlled by finance, service desk people don't trust data from marketing, etc. This situation results in very high costs for maintaining these independent data silos but moreover, it stops people from using all the data that could be essential to them and the business at large. The right tools, like Accurity, can assist with these very issues.

If you have your own company, then there are always monthly data costs but often little underlying knowledge of all that data. So why not use that investment wisely? It is important to bring “context” into business intelligence reports, using the proper data and tools with up-to-date sources. Perhaps what you are doing is not really the best - not enough to have an effective impact. The experience, knowledge, and expertise that Simplity has could be crucial.

We have recently launched multiple, new services to help businesses in this new normal. Industries that can benefit from these new services include e-commerce and retail, finance, insurance, travel, media and entertainment, health, telecoms, education, and energy. If you are interested in any of our services don't hesitate to get in touch.

If you are involved in one of the above industries and you have a really interesting use case that you want to work on with us, why not reach out to us and we could partner up to get things rolling with benefits for both of us such as solutions with reduced rates for you, and interviews/case studies for us.

Petr Mahdalíček
CEO and founder