We spend a lot of time in conversation with prospective and active clients talking about their quest to be (more) data-driven. In those conversations, it can be taken for granted what "data-driven" actually means, and we suspect we'd find quite some variation if we dug in deeper. But that's a topic for another post. Today, in this blog post, we want to investigate how we help you get into a better position to use your data for decision support.
When clients say they want to be data-driven, it's not really the case right now that managers and organization heads are making decisions in the absence of data. Even if the decision is something like "we need to raise prices," there is generally data driving that decision. In this case, the data in question could be nebulous, something like, "we're not making enough money," but there are actual statistics involved. So the idea that we're fighting against "decision by hunch," or that leaders are not making evidence-based decisions, is probably a bit of a straw man argument.
Having said that, it does seem very reasonable to ask about the quality, quantity, and relevance of the data that is consulted, as well as how thorough the examination of evidence is. "Our revenues are not high enough to meet our profit goal" is a legitimate data point, and if that's the extent of your analysis then raising prices could be considered a data-driven course of action.
Even in this simple scenario, of course, there's room for more raw data, more targeted analysis, and a more rigorous review of the evidence. Are competitors selling their product for a higher price? Is there data that suggests existing customers would pay more? Would a higher price point draw other customers out? Are there other ways to meet our profit margin goal? For example, what about selling more product at the same (or even a reduced) price? What about reducing the expenses involved in making and distributing our product? What if we introduced a new product that is inherently more profitable? Or maybe some combination...
What does an organization really need if it wants to be data-driven, if it wants to make evidence-based decisions?
Here's a partial list:
- Data must be sufficiently high-quality data, that is it must be accurate, complete, consistent, structurally valid, and so on.
- Data, whether raw, processed, or fully analyzed, must be delivered to decision-makers so that they have enough time to process it and to include lessons and insights in their deliberations.
- Decision makers, or someone they trust, need to have the data skills needed to understand and apply insights from the data.
- And this kind of intelligence-gathering goes more smoothly when the data is provided in a digestible format.
The data used in decision support not only has to be trustworthy and timely, it also has to be relevant. How do we decide which data is relevant, and how do we determine the best method for presenting that data?
Not to put too fine a point on it, but isn't this really what data governance is all about? In order to apply data (or analytics) to a decision, we need to know what data is available to us, what data is applicable to this situation, what data is incomplete, or inaccessible, or unreliable. And, of course, we also need to know something about which methods of analysis are legitimate and appropriate, and how large an effect we need to observe from the data before we use it to move the needle. For more thoughts on this topic, please consult our previous blog entries on building data literacy competencies.
- When we catalog our data assets to know what kinds of data we're collecting, and where we are storing it, and who is accessing it, we're practicing data governance.
- When our data is documented and monitored throughout its lifecycle, so that we understand at any given point how long we have possessed it, what changes have been made to it, and most importantly what it means to the organization, we're practicing data governance.
- When we identify data stewards, and when we defer to their expertise and actually allow them to act as stewards of our organizational data, we are practicing data governance.
- When we curate data sets and certify data products - that is, when we agree on a single source of truth rather than futilely chasing a single version of the truth - and use those tools to guide our business, we are practicing data governance.
- When we consume data in a spirit of humility about what we do not (and often cannot) know, when we engage with data in order to explore possibilities, when we continually seek to grow our data competencies, we are practicing data governance.
- Finally, when we change our data collection and management practices in order to bring more data to bear on decisions, or to better evaluate the consequences of decisive actions, or in recognition that data quality standards are not strong enough to provide for proper decision support, we are practicing data governance.
Ultimately, then, what are we suggesting? Being "data-driven," or employing "evidence-based decision making," is a worthy and admirable goal, and organizations that fulfill that are likely to have better processes and outcomes. But to use data in this way, organizations need useable data! How do you get there? Start by governing data, as we suggested. Continue by governing data. Need help governing data? We have some ideas. Feel free to contact us.
The Data Cookbook can assist an organization in its data governance, data intelligence, data stewardship and data quality initiatives. IData also has experts that can assist with data governance, reporting, integration and other technology services on an as needed basis. Feel free to contact us and let us know how we can assist.
Photo Credit: StockSnap_SO60NK95EP_groupdecision_managingdata_BP #B1220