At the beginning of the year there are a lot of lists, research papers and opinions what will be the 2019 Business Intelligence trends. When reading these blogs, I find it hard to relate as they are often based around the doings of Silicon Valley giants, who have built their whole business around analytics and data. So, I decided to add my own 0.02 Euros into this discussion.
When I follow behind my own desk from my limited corner view what we are doing in Finnish BI, the connections to these trends are usually sparse. Thus, I decided to take a more pragmatic point of view. My list is closer to earth than the competing writers'.
There are lots of different things I could talk about and a lot had to be left outside. I ended up listing here 2 wishes, 2 guaranteed trends and 2 speculations. And a poem in the end.
1. Master Data Management and Data Curation
MDM is the stone of the data world that we just don’t seem to get moved out of the way. Companies merge, split and restructure. Still they seem to lack ready methodologies for getting the data live and accommodate with companies. It feels that many organizations will go two years backwards in analytics every time some change occurs.
These situations are quite challenging. For example, in merging of two companies the new company must follow
- current state of both mergers separately, so that business can better plan the target state, and
- real time follow-ups must be in place while merging proceeds towards the target state, so that the company can get early warning signals if something has gone wrong in change management.
If you are going to make one promise this year, then promise to put Master Data Management process in place, so that you will no longer generate bad data and the process will take care of all changes. If you do not do this, you can't convince your organization that Business Intelligence can get results. Furthermore, you will not get users to your applications and you can not build any new and shiny Machine learning or AI solutions on top of you Data Warehouse.
2. Digitization of Business Processes
Second wish on my list is that business processes would be digitized and that they would live together with analytics. There are two important points in this:
- Data must be generated while working so that you do not have to do separate savings to manual systems.
- Data and analytics must eliminate more and more repeatable tasks. Just now in who knows how many organization sales people are generating Excel-sheets to their customers to report last year results instead of getting automated reports to both sales and customers from the system with a link to a portal that allows access to deeper analysis.
3. More exposing of private data abuse
Third trend is not in my wish list, but I predict that there will be more exposes of outrageous abuse of our private data. My accusing finger will point here mainly to the governments of the super powers and giants of Silicon Valley. Not in our backyard here in Finland – surely not?
There was last year a lot of talk about GDPR, clarity of data, data governance and ethics. These will be hot topics also this year.
Big law-abiding listed companies are struggling with the new limitations that might be already too tight in some aspects. Meanwhile the bad boys don’t give a damn about them.
What I find surprising is that we consumers have almost totally surrendered to the new reality where our phones listen to us and Facebook sells our private messages. Luckily EU has woken up as my anger alone is unlikely to make the change.
4. Cloud services in Analytics
The benefit of the tightened legislation is that it has forced the Cloud Service providers to get rid of their most ridiculous demands such as ”We own all data that you will put on our platform” and thus it paves road to generalizing the cloud services.
Cloud Services (both PaaS and SaaS) are best possible offering in the analytics side where there is often huge spiky demand of calculation capacity. After the calculations that capacity might not be needed at all for rest of the year as the results are ready for the business to utilize. (For example, in customer segmentation every purchase is compared to each other to create segments from similar purchase types.)
Cloud services will be Finland’s mega-trend this year. There is strong push from vendor side and pull from the IT. Whether it is the best investment to drive Business Intelligence forward or not is another story...
5. Finally, time for User Experience?
There's a lot of talk about usability on different levels. Hype words are Storytelling, Actionable Insights and rise in the adoption of Analytics. The tools have developed greatly here during the last years and it is easier and more cost efficient to develop clear actionable analytics.
The greatest challenge is still the chasm between the technical developers and the operative business people. Developers live and breathe data – data terminology and benefits are clear to them. Inspired by the hype in their bubble they develop fancy models with lots of complex combined analyses to the operative people who are not data people. Business requirements on the other hand are often just to get fast and clear basic information. Most of the users would be happy to get: "How was yesterday compared to January last year."
Many companies have started to develop "MyCompanyData" apps or portals to their customers where the focus is more and more in ease of use and user experience. They are often built on top of licensed software that has good support on mobility, shareability and commenting.
Let’s hope that we will get there also in internal reporting so that we can finally get the top management to use those Management Dashboards. 😉
6. Rise of Machine Learning and Artificial Intelligence?
On top of the current hype curve is the rise of AI and Machine Learning. Here too the tools are progressing fast.
More and more there are (half-)ready algorithm libraries that can be utilized with moderate investments. They have been developed to fewer black boxes so that people can more easily follow the machine deductions. All significant Analytics vendors are investing in this area as their spearheads.
But then comes the famous “but”.
- When data and processes are not in place, experiments usually remain as just experiments.
- Machine Learning and Artificial Intelligence are far from ready, so there might be a high price to pay if you want to be a forerunner, and you might end up funding the product development of those who follow.
- Very, very few companies are built with data in mind.
- Most of the data scientists who handle these techniques are still in the early stages of their careers, so they have not had time to get acquainted with the realities and schedules of the business world.
Thus, the chasm is huge to cross in one year. However, I do recommend everyone who wants to develop their companies towards data driven business to boldly pilot these areas to create new foundations for future growth and to stay ahead of the competitors. In the first pilots I do recommend keeping the scope limited and focus on the road so that you can also finish before the year end.
Poems from the Edge of the Cloud
For years I have been distressed that love is totally over presented in music and poetry while we who are interested in data and analytics are totally uncared for. Thus, I have firmly decided to start fixing this problem. I can’t sing but I will appreciate greatly if someone will someday compose my poems.
Here are Atte’s predictions
Analytic trend reflections.
I hope they did not upset you,
because I forgot what you do.
Instead of giving in to rage,
help us build a better stage.
If you disagree totally,
give feedback openly.