The three types of canvas that matter most

We think Count shines when you use it to close the gap between metrics and the business context they instrument, amplifying their relevance and meaning, and accelerating people’s information gain from them. In short: when it makes the processes, models and assumptions of the business more visible for people. We call this “operational clarity”.

Right, but how does it do that?

While the possibilities granted by the limitless canvas at Count’s heart are, well, limitless. Time and time again, we see customers gain over-sized impact from three distinct models of its usage.

Each of these serves a distinct purpose, but are best understood as connected to the improvement cycle. You might know this by different names from different disciplines, but broadly it captures the way that organisations identify operational challenges, explore solutions, come together to decide, and then systemise the monitoring of progress.

So let’s start there.

Metric map

Obviously, we see Count not as a like-for-like replacement of a traditional BI tool, but as a reinvention of what you’d use them for. If we were to encapsulate this in a single canvas, it would be a metric map.

The goal of metric maps—and the bar to which they must be held in order to remain effective—is to not just create but to maintain “operational clarity” for individuals, teams, and the organization. By this we mean that they:

• Permit people to make decisions and actions with more confidence and conviction, built on a foundation of organization-wide awareness
• Build an ever- and self-deepening awareness of how the organisation works and reacts to change
• Make the invisible in data and context alike, visible to all, and open for investigation by all.

In short, they should help visualise the business, not just the metrics.

A cynical reaction to seeing a metric map might be that it is a “prettier dashboard”, but this reductive judgement ignores that maps aren’t called “prettier directions”.

Firstly, maps encode huge amounts of information, but in a way that at its heart, only requires you to follow a line between two points, but forces and encourages aware of the scenery along the way.

Secondly, while a map may feel like absolute precision, in usage we experience them as a series of relationships: My next turn is the third on the right. There will be a pond half-way before that. This area looks busier than that.

That metric maps can bring the vivid customer experience of an onboarding process together with the measures of its performance is only half of the picture. The relationships and dependancies between sets of metrics, between options on an application screen, or spanning qualitative feedback, a touch point, and a forecast of growth are what incite people to understand, and excites people to explore.

Collaborative canvas

Where the canvas shines is breaking through the limitations of notebooks and other analysis environment and bringing people around the same information to build understanding together.

Data notebooks went quite some way in bringing literate programming to the mainstream data world; connecting code, colours and corollary, and helping viewers understand analysis and insight . While they are excellent tools for linear explanation, they are often secondary and supplement to non-linear exploration.

Canvases allow for Python, SQL, no-code and tabular cells to be composed together, showing dependencies and connections, and allowing analysis to be laid out more freely. Detours can be taken from the main thread, with more flexibility for discussion and documentation.

By making these canvases collaborative, analysts can involve others across the organization in their work, gaining perspective and insight, and beginning the communication process in the midst of analysis, rather than at its end.

Collabrative Canvases also invite new ways for data teams to work with the stakeholders: where one can lay out the questions, assumptions, and illustrations while the other assembles the data, confirms hypothesis, and identifies ways of monitoring.

Explorations

The last type is not a canvas at all; it is the canvas you don’t make. We want data to be accessible in every decision, and drive every best decision, but sometimes a decision is purely operational and ultimately ephemeral.

What does this mean? What should I do next?

We believe it should be easy for anyone in an organization to see a collaborative canvas or metric map, and then “zoom in” on a single point.

Good visualisations should be explorable explanations: snapshots of interpretation that invite people to poke around and grow their understanding or answer a question that it provoked. But what segments make up this sales number? Is this seasonal? Where should I focus my attention this week?

That’s why we’ve made it easy for cells to be explored, and for anyone to use governed metrics to build their understanding of what is being shown and how it relates to their moment in time perspective.

Explorations can be fleeting. Enough to serve the operational decision and discarded after use. Other times they can be the seed for a canvas of their own; a chance to capture what is being seen by one, and involve the thoughts of many.

It is in this way that these three types of canvases are distinct but also cyclic. A metric map may alert us to something requiring investigation through a collaborative canvas, or inspire someone to explore a particularly element. Either of those acts should have the freedom to challenge and inform what that map is showing, and by doing so enhance the clarity the organization has.

At Count, we share our belief that dashboards are both the symptom of a poorly-functioning data-culture with all that listen. They start with the noble model of a car dashboard: a slimmed down and focused view of information, set within the context of the world outside the window, enabling smarter decisions within it.

The reality for many however is more reflective of the earlier etymology referring to the wooden board positioned to present mud and debris being “dashed up” into the faces of passengers of horse-drawn carts.

We’d argue that modern dashboards rarely live up even to this. They obscure the view of reality, box you in, and, well, further separate those riding in the cart from those trudging through the mud.

Instead, we think the three models of canvases presented here steer organizations to a healthier relationship with their data, and between data and functional teams.