Demystifying the Single Version of Truth

Larry Seligman, VP, Advanced Consumer Analytics, InterContinental Hotels Group
100
144
27
Larry Seligman, VP, Advanced Consumer Analytics, InterContinental Hotels Group

Larry Seligman, VP, Advanced Consumer Analytics, InterContinental Hotels Group

Introduction

“Single version of truth (SVOT)” is an equivocal term that I’ve often heard described as desirable but rarely heard explained. In the context of analytics and insights, it’s often used to describe how key stakeholders only want to see one set of performance numbers or one explanation about what is driving performance. Unless an organization has clear reporting and analytics accountabilities that are widely accepted and enforced, multiple versions of the truth can make their way to senior leadership, or even to shareholders and key clients.

In big data environments, as the number of data sources increases, so does the potential to create multiple versions of the truth (MVOTs). Data integration tools, visualization platforms, and reporting tools are making it increasingly easy to create new versions of the truth. When there are proper controls in place, those versions can be minimized, but many organizations don’t have that governance structure in place.

Truth: Single Version, Multiple Levels

But what does it mean to have an SVOT? What truth are we singularizing? We can think of our business “truth” in five levels:

Data – The simple raw transactional detail. Who bought what, when, and where, and who sold it to them. Any basic transactional facts.

Metrics – Sales per month, revenue per store, customer satisfaction, anything else to be measured.

Reports and visualizations – The presentation of collections of metrics. Performance by channel, customer satisfaction by store, top sales performers. Generally, the presentation of metrics and data.

Analysis – Why are our sales up? Why is our customer satisfaction down? What are the drivers of financial performance? How does overall experience relate to financials? Analysis can also include predictions (will we hit our targets) and observations (same store performance is up for the fifth consecutive quarter).

Narrative – What is important to know? What decisions do we need to make? What needs our attention, and what does not? This is the critical piece that frames the other levels of truth.

  ​Data integration tools, visualization platforms, and reporting tools are making it increasingly easy to create new versions of the truth 

There are dependencies among the levels: metrics depend on data, reports and analysis depend on metrics, and data, narrative depends on reports and analysis. If there are MVOTs upstream, there will be MVOTs downstream as well. If you find analysts producing contradictory models and findings, then the misalignment could be in the analysis or at any level under it.

Dimensions of the truth: spanning time and space

Alignment at these five levels many not solve all of your SVOT problems. Beyond just identifying the right source of data, if historical transactions are rewritten over time as part of a data quality effort or for other reasons, the data pull done today may not be the same as the same data pull for the same dates done next month, or at year end. So, the permanence of your data will impact the permanence of your truth.

Truth can change as new data, metrics and analyses are done. Strong communications are critical to minimizing the persistence of antiquated truth. If a meme regarding the value of a digital impression has been circulating throughout the company for years, a new analysis showing a different finding will not change what shows up in business cases and value stories until it has been properly communicated. Even then, adoption of the new truth is not guaranteed. Senior leadership must understand and adopt the new truth, and insist it be used in forthcoming deliverables; otherwise, it may end up as just an alternative truth.

Truth can have specific use cases as well. A customer segmentation that results in five customer segments need not serve every purpose. So the truth that “we have five customer segments” can be over generalized, in which more specificity about the purpose of the segmentation is needed: “For private campaign targeting purposes, we have five customer segments.”

Ownership of the truth

Ownership of these levels of truth need not come from a centralized data and analytics function or center of excellence. Clear accountabilities and decision rights are more important than organizational structure for reducing the versions of the truth. Those accountabilities and decision rights must be recognized by the executive committee in order to achieve an enterprise SVOT; otherwise, alignment of the five levels of truth may only result in a functional SVOT. A certification process can be helpful for distinguishing commissioned truth from truth that comes from a rogue team.

In my experience, many people want to own the development of the analysis and narrative levels, because they can position the story in a way that best serves their purposes. Fewer people are interested in owning (or funding) the lower, more foundational levels. Blame tends to get pushed downward: the data is faulty, the metrics are looking at the wrong thing, or the numbers were pulled off the wrong report.

Can your organization handle the truth?

Not everyone in an organization desires a SVOT. There will be some purists who desire no equivocality, and some pragmatists who believe that things will all be much easier if everyone is on the same page. But some people like the MVOTs because they have learned to navigate them. Others philosophically agree with the idea of having multiple viewpoints that allow them to triangulate on a common understanding, rather than having a SVOT dictated to them. Whether it’s for pragmatic, purist, political or philosophical reasons, some resistance to an SVOT should be expected.

So if your organization is pursuing a SVOT, pursue it when it is the only way to solve a problem, or where it delivers clear financial benefit, or is mandated by the highest levels of leadership. Make the scope of your SVOT project meet just what is necessary. Follow good practices of gaining alignment to avoid a power struggle at the end. Build on small wins. Communicate early and often. Then, in those areas where it is most critical, the truth can set your organization free.

Read Also

Getting the Most out of Big Data

Getting the Most out of Big Data

Kris Rowley, Chief Data Officer, GSA
Big Data: Separating the Hype from Reality in Corporate Culture

Big Data: Separating the Hype from Reality in Corporate Culture

Brett MacLaren, VP, Enterprise Analytics, Sharp HealthCare
Maintaining Maximum Relevancy for Buyers and Sellers

Maintaining Maximum Relevancy for Buyers and Sellers

Zoher Karu, Vice President and Chief Data Officer, eBay
Building Levies to Manage Data Flood

Building Levies to Manage Data Flood

Adam Bowen, World Wild Lead of Innovation, Delphix