Digital marketing and data analysis are inseparable from each other. Marketers are also required to have the ability to analyze data year by year.
As a result, I participated in a data analysis study session called Marunouchi Analytics to obtain information.
The study session consisted of two parts, LT + panel discussion. I will write down what I thought while summarizing the contents of the discussion.
The following data scientist greats were members of the panel discussion.
- Mr. Hirao, Saika Co., Ltd.
- ABeam Consulting Co., Ltd. Mr. Homma
- Mr. Kamiya, Rinoshisu Co., Ltd.
- Recruit Lifestyle Co., Ltd. Mr. Harada
Everyone is too keen on AI. Isn’t it more important than that?
I often hear about AI that automatically analyzes data only for some functions such as web analysis, but how do cutting-edge data scientists perceive AI?
The session is“Everyone is too keen on AI. They don’t drink, right?”It started with Mr. Homma’s unique story. This is what followed after that.
- You can’t see the surprise with only AI and deep learning
- Innovation is in the unexpected part
- Surprisingness is often seen from the data part
Hmmmm … If you just want to do business within the current framework, AI may be an alternative. However, it is impossible for AI to make major improvements or make new discoveries beyond the current framework. That’s right.
And to find “surprise”Communication between scientistsIs also important. Therefore, rather than paying a high cost for introducing Watson etc., it is better to hire excellent scientists and communicate while drinking. I see, the story of the first sake continued here.
“We have to think of ideas that cannot be created with AI.”
The words that Mr. Homma said pierced me.
Business side and scientists need to compromise
Recently, the tools have evolved so that you can see the data without much specialized knowledge. I think this is accelerating the flow of “Let’s analyze data for marketers too!”
In fact, it was unanimously recognized by all panelists that “the hurdles for working with data are getting lower year by year, and the tools themselves can be touched without specialized knowledge.”
However,View data → A big wall between using dataIn the discussion, there was an opinion that it is dangerous to make a judgment from data without specialized knowledge.
Even if the hurdle is lowered, it is a rough story to leave it to “I can do it because I can do it!”, And there are still many parts such as how to view and handle the data.Data scientists need to approachI wonder.
Judgment based on data seems to be persuasive, but how to take the original data and how to interpret the data that came out,Is that really correct?If you don’t hold it down, there is a possibility that you will not be able to demo.
The key to data analysis in the future may be how data scientists should be close to the business side, create an environment for using data, and manage the business side regarding the correct use of data.
Do you really need state-of-the-art analytical technology?
While there was a talk that “even if tools become commoditized, data scientists will not disappear. The way of working will change little by little,” Mr. Kamiya said that the way of working of data scientists will change significantly. was.
Mr. Kamiya is in the data analysis consulting business, but to the data scientists“I don’t need half of you”It seems that you are saying that.
The tools themselves are getting smarter, so analysis that would otherwise be impossible without a data scientist can now be completed with just the tools.
Of course, advanced analysis requires a high degree of expertise, which requires a data scientist. However, the data analysis methods used in the actual field are certainly limited, andThere aren’t many cases where a data scientist is really neededIt may be …
Is it a scene that requires a higher degree of specialization in the future, or what data will be collected? What to analyze? It may move to the upstream process as a job.
What are the necessary abilities in the future?
There have been many talks, but what about the skills you will need to analyze data in the future?
The following skills were mentioned in the discussion.
- Ability to find variability
- Ability to not read the air.Ability to see what is not seen
It was the latter that I thought was interesting. There are certainly areas where no one wants to touch, even though they have obvious challenges … I felt once again that putting a scalpel in that part from the perspective of data is also an important skill.
If you put a scalpel in a place where there is no data or it is not converted to data in the first place, you will surely get results.
Bit A Kawaguchi was also on stage
Our data scientist Kawaguchi was also on stage at LT. He is also one of the members who host Marunouchi Analytics.
Marunouchi Analytics is
- Practical-based analysis study session, case sharing session
- There is a low threshold for entry from other industries and occupations and a wide range of themes.
- A place that leads to work
It is an organization that holds study sessions with the importance of. We are holding other events, so please check them out.
>> Marunouchi Analytics Official Website
Nowadays, not only the web but also the real people are getting more and more data.
“It’s impossible to analyze data, just leave it to someone who can.”
I think it would be ideal for marketers and business planners who know the business best to be able to analyze data, as the times like this are probably over.
However, throwing a ball saying “I have a tool!” Is not good, and if you can analyze while talking well with a data scientist, you may be able to produce even higher results.
See you soon.