Making decisions is challenging, even in the best of times.
Whether it’s a monumental decision with dire consequences (e.g. when to reopen the economy at the expense of lives vs. livelihoods?), a trivial decision (e.g. should I watch Tiger King?) or a business problem (e.g. given the COVID-19 slowdown, where should we invest our sales and marketing efforts?), it’s difficult to solve problems.
One part of the complexity is information. Too much vs. too little.
We think more is better (and it is when you have good quality information). But in reality, it’s often an easier problem to solve for when you have less information.
Wait, why would less information be better than too much?
There can be clarity in a lack of data. Here, the business problem may be well understood, but the information to solve the problem is lacking, with everyone agreeing that vital data is missing. You have a missing data problem.
How do you solve a missing data problem?
You collect it. You implement systems to solve the data problem. This then helps you solve the business problem.
Often smaller organisations know that they can’t fly a plane without vital metrics and instruments to measure performance. They stay focused on collecting the information and putting systems in place to ‘fly the plane’. They lead based on the signals that they have, striving for more clarity, insecure that they don’t have access to the so-called wealth of insight that larger organisations have access to.
In contrast, large organisations drown in information. They have access to spreadsheets, dashboards, databases, tools, insights / BI / analytics teams, syndicated reports, publications, consultants and research agencies. The quality and context of the information gets lost in ‘user permissions’, namely what you have access to.
From a CFO’s perspective: “we’ve already got access to Excel and Tableau to service the business with insights and metrics”. Nobody wants to write a check for more data or new systems, particularly in a recession.
But ask 100 people in the business: Who are your top 10 customers? By region? Which of your top customers have stopped spending money due to COVID-19? What is the lifetime value of these customers? Who are our top 10 suppliers by spend? How profitable are our core products/services?
You’ll get 100 different answers.
Instead of factual data-driven answers, you get qualitative responses. Stories. Answers that depend on who you ask, the systems they use, stories that describe how comfortable they are with data and the tools that house said-insights.
This is not COVID-19 where quality data, or a lack-there-of, doesn’t exist without adequate collection and testing (e.g. population-based, age-stratified sampling) to understand the magnitude of the problem.
Nor are these tricky questions, nor depend on complicated calculations or predictions about the future. And to cite the CFO, “the data exists”.
So if most organisations don’t have a missing data problem - with the information residing in databases, spreadsheets and BI tools - what problem is this?
This suggests that there is either a data quality issue or a fundamental problem with the systems - the spreadsheets or dashboards - that communicate the metrics to the business.
It could be a data quality issue. It’s unconvincing, however, that a Fortune 500 or ASX 100 company with advanced analytics, data science and data engineering capabilities can’t or haven’t solved this issue for key business metrics and datasets.
This suggests that the problem is likely a systems issue. Namely, that the current tools - the spreadsheets and reports - are ill equipped to give teams factual, data-driven, answers to simple questions such as: ‘Who are our top 10 suppliers by spend?’.
Note, you don’t have to work in data to spot a systems issue.
As I am writing this article, I received a call from my ex-car insurance company asking me whether I’d like to renew my contract for roadside assistance (e.g. breakdowns, flat batteries etc). The only problem is that this is three-months after the bill was due. For me, it was intentional to not pay the bill. I sold the car so no longer needed the service. But for the organisation, contacting me so late after the renewal date suggests a mismatch between the facts (i.e. the raw data saying that the contract is up for renewal in February 2020) and the action (i.e. the system prompting the team to contact the customer).
So how do you solve a systems issue?
You can do one of three things: (1) do nothing, (2) use the same systems differently, or (3) implement new systems designed to solve the problem.
There is a cost to each of these actions.
Doing nothing relying or tweaking current systems results in ineffective decision making.
Like asking 100 people for data-driven answers and receiving 100 different answers, this is akin to flying a plane without functioning instruments to navigate calmly and confidently. Try and make sense of the data you have, cross your fingers and hope for the best.
If you’re a data whizz or in a finance team, the current systems likely make sense to you… and that’s ok if your entire company employs only data-savvy people (highly unlikely).
For example, to a finance team, a spreadsheet is more than a bunch of numbers. It’s a story about the health of the organisation.
But to the majority, to the other 9900+ people in the organisation, it’s just a bunch of numbers that are difficult to interpret and don’t tell a story in a way that is meaningful to them.
The very fact that this problem exists, means that current tools have not solved the systems problem of scaling insights and the capability to interpret metrics across an organisation of 10K+ people, comfortable in their legacy ways.
Using the same spreadsheets and reports albeit trying to do it differently, means training your people to somehow get to a place where they can use data to answer simple business questions like ’ Who are your top 10 customers?’. But this doesn’t scale across teams, data, and puts a lot of pressure on people to become more data literate.
In contrast, systems such as Hyper Anna are designed to solve the very problem of giving people, regardless of their background, data-driven insights to make decisions, such that if you asked anyone in the organisation “Who are our top 10 suppliers by spend?” they could answer this question instantly. Reliably. Confidently.
Don’t ignore the systems problem.
It doesn’t matter how much data-prep or predictive analytics you do, your organisation cannot make data-driven decisions at scale, without investing in systems that automate the process of giving your entire organisation, not just data-savvy teams - clear, factual insights.
And the cost of doing nothing?
This means flying blind, particularly in a recession where you have less headcount, less revenue and the pressure to cut costs to protect margins is unprecedented.
Given COVID-19 and the livelihood of both human- and business- survival, there’s never been a more critical time to truly evaluate how you and your organisation makes decisions. We expect our world-leaders to do so, and thus, should also act in ways consistent with the values of data-driven/evidence lead decision making.
We urge you to think strategically about the systems you have in place to report on metrics.
Don’t leave it to luck.