DATA - OUR RULER, OUR RELIGION, OUR SAVIOUR
Updated on October 7, 2020
A blog post written by our very own Ágnes Zuzan - originally in Hungarian 🇭🇺 , about problems she has noticed while working on many different data digitalization and transformation projects.
Picture credit: StockFreeImages
"By any chance, have you joined a cult recently?" - asked my husband cautiously, leaning into the doorway of my work-at-home lair.
I haven't even yet bothered turning back - my mind still miles away, mesmerized by the rows and rows of numbers on my laptop's screen - while he continued, laughing:
"All I heard during the last 30 minutes was DATA DATA DATA - you just kept chanting it endlessly!"
I opened my mouth for a witty retort, but nothing come out as the realization dawned on me - indeed, data is our religion. It intertwines our whole life, it is present in our main decisions, and consequently we are leaving gigantic data-footprints all over behind us.
But how are we utilizing it? Are we slaving for “all ruling data” by endless recording, tracking, manually adjusting and misinterpretending, or can it be our saviour?
I strongly believe it is the latter - we should turn data into our advantage and use it to provide such insights and predictions which were never ever available before - and which can drive our business as nothing did it before, no matter what industry or sector are we speaking about.
But even so, to make data work for us – and avoid us slaving away endless hours of manual labour to get the perfect data – there is still a long way to go.
While working on several different data digitalization and transformation projects, I kept noticing the same issues:
- Quality – manual input issues: data is there, yay! Wait a minute – why is it being manually adjusted randomly? And how to reflect on this fact by our all-round automated shiny process? Oh well...
- Quality – root cause issue: as painful as it sounds, it worth considering rooting the issue out completely, starting from the lowest level, and realizing the back-end / data cube is simply no longer fit to store our data in a format and functionality which would suit today’s requirements.
- Availability: we do not know what data our organization / corporation has. Well, there are assumptions. Which are mostly optimistic, following the line “it must be fairly straightforward to source the data” and which nearly always end in complicated data availability checks and in bitter disappointment.
- Taxonomy: we are not speaking the same language, and it shows – in data quality. A product or service for example can mean a whole different bundle within different subsidiaries at the same company, or sometimes even within the same company, in different systems.
- Naming: similar to taxonomy but a bit different topic, naming conventions can make all the difference and should be considered before integrating data.
- Readiness: although the data is confirmed to be existing and being available, it is a whole new story to feed into the system which will further proceed / visualize it. Most likely figuring out the possibility of real-time or scheduled data refresh is still ahead of us.
- Feasibility: we do have the data, it is ready to use, but is it feasible to use for the purpose we want to? Can it be merged in a meaningful way? Or switching the question the other way around: will the most requested visualization really makes sense, or we just need to keep it in scope as must-have, due to internal politics / old habits?
- The “hidden value” factor: we don’t know what treasures are we sitting on – it might happen that although the company is completely clear on what is available overall, the insight what can be achieved by the combination of the data set has never occurred to anybody, as it is lost in the sea of data.
- Security: at several digital project, the biggest blocker can be IT security / GDPR: data cannot be shared, stored, used, and visualized, without being anonymized, encrypted and necessary accesses set. These are of course a must if we want to keep our data safe, but being typically time-consuming activities, it is worth a careful planning.
- Internal company politics: the data may exist, but the ownership is with another team. To gain access, x+n internal approval and decision necessary which can take up valuable time, which the project team would rather spend on integration and implementation.
- Variety of possible sources, a.k.a being lost in the corporate wilderness: all happy, we are using the right data set, real-time no less! And then the realization hits, that there is another system hidden in the corporate matrix, which is even better as a data source, but has a completely different owner, back end, hierarchy and interface need. You guessed correctly; we are back to square one.
- Transformation issues: we have our beautiful, perfect dashboard / system / app – and everybody keeps demanding xls extracts and/or getting the legacy products back. This is a very frequent problem and can cover different underlying issues, from the need of cultural changes across lacking training or handover, till the fact that the visualization might be built without considering user stories / user experience or based on outdated requirements. Being set in a traditional way of thinking can also impact the start of a data digitalization / transformation project, if the involved parties cannot imagine what the end result can bring as benefit, hence not able to draw up requirements/user stories accordingly.
- The pain of parallel operation: nearly always there is the dilemma of the legacy system(s) or process(es) – whichever is the case, we always should think carefully in advance about setting up processes, accesses and protocol for handling the transition period.
- Disappointing end results: we have our new tool / report / dashboard / app, but no improvement in business drivers. Why is that? It can happen business priorities were never really captured properly. Or they were compromised along the way. Or the whole corporate strategy shifted during implementation – you name it, it all can happen. This might be the trickiest issue of all to address, especially that sometimes there is no real issue behind it really, only high expectations: viewing data analytics as a magic wand to all problems.
Colours of Data is a consultancy where data is the new gut feeling.
We are focusing on identifying and rooting out these issues by working closely with our clients, getting to know their organizational and technical setup, and most importantly by understanding as early as possible what they really need to achieve and how we can make their goals happen.
Along with enabling the best set of data, we are making sure to present in a way which provides an insight which was never available before.
The key behind turning data into our advantage - and let it drive our business like it never did before - is to carefully identify the goals we want to target, and the problems we need to face.
Keep your eyes on our social media channel for our next use cases - we have just set foot in new markets to help clients in different industries to achieve breakthrough data insights which will strongly boost several business drivers.