Although we are emphasizing data analysing the value that begins professional work to the enterprise all the time, but the much denaturation as a result of market environment and company business, we are very difficult fore-and-aft assessment the value that data analysis brings to the enterprise, this also makes a lot of enterprises analyse platform existence misgive to whether wanting deploy data.
And from the horizontal stroke up, the value that data analyses is proved more easily. The near future, an analytic orgnaization of foreign to the past thousands of realized gain business telephone call inside 5 years undertook dogging, discovery is in these phones, the term related to data appears most is " data is analysed " , its about 5 from 2014, 000 grow 16 2018, 000, what this proves the value that data analyses got these businesses is general approbate. Directer evidence is, these classics regular meeting speak of the company that data analyses, income increase rate compares tower above of person of the same trade 12% .
A trend of world of this since business, also be the data expert judgement to future: No matter whether we like, algorithmic and decision-making coming, that is to say, more decision-making it is by the data of automation the analysis comes of drive, the subjective judgement of mankind of and rather than.
It is all-purpose that data analysis is not,
But, what we still should notice is, as a result of the complexity of real world, even if the data that the enterprise used high quality, and firm data analyses a process, data analysis result is very scientific also close, but the result that still can bring expect to be less than.
It is with American mainland airline exemple, in a few years before, its discovered fuel is one of highest cost that its industry runs, to reduce the fuel in flight process use up, its made the index that tracks specific fuel consumption, air man of shunt excitation encourage uses fewer fuel. Below the drive of this policy, these air man begin more use slow fast flight, reduce the kind such as air conditioning temperature, cogent let what specific fuel consumption got fixed rate reduce.
But, this kind looks complete the result that got expect to be less than however by the company policy of data drive, to the passenger, of flight speed reduce can bring about time-consuming addition, reduce air conditioning temperature to be able to affect easy directly measurable. Accordingly, after this policy is carried out, the passenger's satisfaction is spent had reduce significantly, affected the accrual of airline and brand directly.
In home, we can discover a lot of these case likewise. A simple case is, a lot of companies established rigid KPI system to undertake assessing to employee (the visit quantity that is like an user, change the target such as rate) . This can promote employee the initiative that implementing short-term goal course admittedly, but from will for a long time look, also may cause a company too nearsightedness, those who ignored the maintenance to user relation, brand is long-term model, and the social responsibility that should bear. These problems can be not reflected in data index, but bring about an enterprise very easily however to accumulate risk element gradually, bring the serious consequence that cannot expect finally.
These case inspire us, when the organization needs to collecting new data to found the index that provides insight more, return the negative effect that needs more reflection to may be brought. A cleverer way is to have small-scale test, so that be in,this program will check to whether be put in unforeseen consequence before the plan popularizes whole organization.
Donald Feinberg of Gartner executive analyst was offerred the following 3 principles, will help an organization answer in the process to gradual progress of data drive organization, what need faced challenge and hook:
● often reviews the data drive strategy that the organization formulates
The unforeseen consequence that ● produces to data index place undertakes thinking adequately
● tries new data, new measure and new analysis model
Intelligence of labour of choose and employ persons is decision-making feasible
From above discuss our already OK and clear discovery, decision-making it is actually one be quantified very hard, also be a kind of very complex operation, current artificial intelligence technology lacks this kind of capacity, its reason depends on his bringing into all sorts of elements very hard the analysis not only in the category, also depend on artificial intelligence decision-making once improper, the distrust that can bring everybody more easily (for example, drive automatically if the car produced an accident, the condemn voice that its bring and doubt want Yuan Chao to drive artificially too accident) .
Although we are inside very long period of time,still should depend on artificial and decision-making, but we also need to realise, data analyses as decision-making as automation value to depend on, its were offerred increase model ability props up us to undertake more deep have an insight into.
The basis is forecasted, to 2021, the author of digital chemical industry of 25 ％ can use virtual personnel assistant everyday. To the organization of those ready-made that it is AI, they need to realise, algorithmic and decision-making from long-term will look is inevitable, the enterprise should be dedicated replace at using artificial intelligence to enhance decision-making, and rather than, produce artificial intelligence and mankind's respective advantage.
- data analysis is revealed use DataHunter-
