Trusting your data and trusting analytics results are separate but related topics. Your analytics initiative can positively bridge both, especially if your analytics solution factors data quality into its results. It is not advisable to delay analytics efforts until an ideal state of data readiness is attained, since organizations seldom reach that state due to the constantly changing nature of their data.
There are several approaches to developing user trust in analytics results. The first is to choose use cases where results can easily be validated. Early in your analytics program this method can help all end-users and stakeholders become comfortable and confident in the use of data-driven decisions based on analytics results and recommendations. This method can also help to detect data quality problems to correct the problems, and train the analytics to recognize and discard “bad data.”
Field service efficiency - Sending properly qualified and equipped field personnel to a customer site or the location of a remote asset is costly. If personnel make an unnecessary trip or need to make repeat trips due to lack of proper information, scheduling, tools or parts, costs quickly double or triple. Analytics can ensure that tasks assigned to different functional areas are all fixed on the first visit, that schedules and routes are efficient, and that postponed work won’t adversely impact quality, safety or revenue.
Improved capital efficiency - Spending less money, or getting more for money spent, is clearly in every organization’s interest. Analytics can pinpoint whether to repair, refurbish or replace strategic assets, which capital projects will have the greatest impact or return, and which projects can be postponed or cancelled without adverse effects. The larger an organization’s budget, the more dollars can be saved by a one or two percent increase in capital efficiency.
Risk reduction – Organizations of all types have varied risks that they must manage from availability of supply to market demand. Analytics can provide value beyond simply identifying exposure to show how to best deploy funds, assets and staff time to mitigate risk by modeling probable outcomes and prescribing the best course of action in any situation. Other organizations, insurance carriers specifically, specialize in managing exposure and risk. In this latter case, effectively managing risk directly impacts a carriers own costs for insurance that in turn impact quoted premiums. Modeling possible losses is valuable for regulatory documentation and compliance, along with various safety, restoration and legal costs when losses occur.
Question 3: Can We Trust Our Data and Your Analysis?
You might be reluctant to embark on an analytics initiative because the quality of your data is unknown. After all, good data quality is important for accurate, reliable and actionable results. Some analytics solutions can help by recognizing and disregarding anomalous or redundant data and flag suspect data for cleanup. Analytics that possesses such capabilities makes it possible to move forward with an analytics implementation, and in the process of that implementation, proactively identify data quality problems that need to be addressed.
In some instances, a score may be used to convey to users the confidence of the analysis. Even though an analysis may show a positive result, the confidence in that result may be low due to a small number of data points, old data or incomplete data. In other words, a solution is more trustworthy and valuable if it includes and conveys a confidence metric along with results and recommendations based on the quality of the input data. Results and recommendations derived from poor data quality will have a low confidence score, whereas results and recommendations derived from good data quality will have a high confidence score.
Each source of data presents its own challenges related to data quality, as does the matter of streaming data “in motion.” Data at rest can be evaluated by analytics to help identify and correct data quality problems. Data in motion, such as from devices on the Internet of Things, poses additional challenges since analysis (and corresponding decisions) might be needed before data is stored. In these instances, analytics performs a critical role of cleansing and correcting data that might be collected out of sequence, reported multiple times, captured inconsistently and so on. Without this analysis of streaming data, it would be difficult if not impossible to present accurate information and alarms to users.
Choosing an Analytics Provider
When choosing an analytics provider, it is important to have an understanding of your business requirements that will drive the determination of what you need analytics to deliver for your organization. It will also be imperative to ensure analytics solutions are able to deliver on the business benefits and results you are seeking, find analytics vendors that understand data quality and ensure analytics solutions can interface with data across your organization in terms of its disparities, volume and frequency of delivery. Only consider analytics solutions that can present the data in intuitive visual formats that simplify the decision-making process and draw attention to important information and look for solutions that operationalize analytics allowing users to interact with analytical models, their outputs and the underlying data in different formats.