Snowflake Software is now part of Cirium
Snowflake Software customers have access to the largest, most complete timely aviation data and analysis.


Questions about the acquisition?
Contact us.
Request or Question

7 steps to building a data-driven decision culture

Team-wide digital fluency results in better and more efficient decisions

Securing data is often the easiest step on the path to a fully data-driven organization. With more than 3,000 terabytes of information, it is likely Cirium has what an organization needs, or can work to help organizations make better use of the data already on hand. The difficult part is creating a culture where data is a standard part of operations and naturally incorporated into strategic decision making.

Teams incorporating data and analytics into daily work product, perform with less conflict and make decisions faster. Teams don’t need to be data dependent, but a data-driven operating framework can validate instinct and grow tribal knowledge.

Peter Drucker said “culture eats strategy for breakfast,” and he’s right. A data-driven culture is a modern answer to better marketing practices. A culture of data drives the innovation and iteration that leads to positive outcomes. Below are key steps to get teams and organization on to the right path.

1) Set KPIs. Set targets for all projects and carefully choose what to measure in advance. Focus on metrics you know which can be acted on. Include indicators to help define audiences to score the impact to business priorities and measure the overall effectiveness of programs. If a program is new or doesn’t have specific goals, use KPIs to create benchmarks for future evaluation. Establishing KPIs gives you material for nonjudgmental conversations about performance.

Leading organizations work to build shared goals and measure shared outcomes through benchmarking . Many companies in travel track program performance,  but without context or competitive data, it is only half the story. When KPIs are designed to incorporate both internal and external factors into one formula, teams can start to get a real sense of how to act to change the data in the company’s favor.

2) Forecast results. Forecasting can be difficult at first, but the process of projecting how a tactic will perform provides a learning opportunity every time. Forecasting makes results analysis easier by providing a pivot point for discussion, and establishes a background for improving KPI selection during each iteration. Have some fun with it, challenge your team and reward the person with the most accurate prediction.

Forecasting in the current climate of constant change can be hard. Our teams are working daily to produce a variety of models and insights for scenario planning.

3) Always develop hypotheses. Where forecasts are about broad performance results, hypotheses are focused on learnings surrounding a single topic that, ideally, can be isolated from other variables. The best way to test a hypothesis is with a simple A/B test. Topics can often flow from creative debates on things like which changes are most visible to the traveler or whether an operational change that seems small internally carries more weight with certain consumer groups.  Treat hypotheses as part of the overall planning and focus on something you hope to learn from the data each week or month. Then test or observe.

4) Integrate a learning plan into reporting cycles. Hypotheses propose an answer to a single question. A learning plan strings together a series of hypotheses to answer a series of related questions to further refine insights and tactics. By incorporating a learning plan into planning, structure and prioritization around hypotheses can be created. This enables you to look at a single topic using multiple variables, from a variety of datasets. It makes sure ensure learning something from every piece of data and it helps to capture that learning for future application.

Finding the hidden stories in the data is how Cirium’s data experts provide unique value to customers that need assistance building data fluency.

5) Putting data in slides is not enough. Set up regular discussions to review data and encourage questions and curiosity through collaborative analysis. Discussion and debate lead to new ideas and test concepts.

6) Free the data. Provide access to anyone touching operations, from a project manager to a guest services specialist. By providing the entire team with access to data and facilitating conversations about it, you’ll see understanding, acceptance and trust quickly grow. Adjustments to resource planning, account management and more will begin working their way into programs organically.

Data visualization is a great step to both freeing the data itself, and making the insights more available for discussion.

7) Don’t be a robot. AI is a key stepping stone to deeper, more targeted insights, but business decisions are made and consensus is formed by humans. Don’t make binary decisions based on data; let data inform and influence the business owners and frontline customer experience leaders.

For more ideas on how performance analytics can be brought into your business, listen to a recent recorded discussion between Cirium’s Paul Trewin, and Seera Group VP of Data Louise Blake.

For further information, connect with the Cirium team.

Find out more about how our capabilities can benefit you

Explore solutions