How to Design a Data Project for Your Business

Looking to embark on a data project in your business? Here’s what you need to consider before you get started.

Creating an advanced analytics program begins with asking the right questions; which business problem am I trying to solve and which data points can help me solve it? Making an error in this initial phase is what causes most data projects to fail.

It’s no surprise that data has jumped to the top of many leaders’ priority list. The need to modernize and optimize is increasingly prevalent through the rapid growth of industries such as platform data solutions, machine learning, IoT, serverless computing, data management and similar fields. At a minimum, many have made moves like implementing a shared CRM tool to manage leads and deepen customer relationships, replacing individual notes with a shared database, and using personalized marketing via integration and automation.

However, a successful data initiative offers far more opportunities than most realize. Data can deliver competitive advantages, develop operational efficiencies, enhance offerings, innovate revenue models, enhance customer experience, and improve sales and marketing efforts. Companies leverage data to streamline operations and decrease operating costs, therefore amplifying profit margins through the digitalization of solutions that create preferred delivery mechanisms by their customer base yet boost profitability by cutting production, distribution and service costs.

Data-driven decisions can help identify new markets plus grow underserved/niche verticals that companies can dominate when positioned correctly. Examples include the agricultural industry, where farmers are honing data and analytics to better predict weather patterns that help forecast the grade/quality of the crop.

Leaders must understand that a successful data project requires diligent effort upfront to identify and prioritize opportunities and/or problems, build the right team, ensure processes and reporting structure aligns with the company’s goals, and more. In a recent survey, 77% of Fortune 1000 executives reported challenges adopting big data and AI initiativesOpens a new window . This number is up from 65% in 2018, showing a significant increase in the realization around the challenge implementing data initiatives.

These problems continually grow despite significant investment – 55% of companies surveyed spend over $50M, and 21% over $500M, on big data and AI initiatives! Further, respondents say that this isn’t a technology problem – 95% of the reported issues were related to people first and processes second.

Building a Data Literate Company

A successful data project takes more than enthusiasm and budget. It requires a “data literate” company, meaning that the culture is built around data and using it effectively. The data literate company has transitioned from the traditional “knowing culture” (making decisions from heuristics and best-guess anecdotes) to a “learning culture” that makes objective, data-driven decisions.

This move to being data literate starts with executive support via a “top-down” approach, then moving into every level of the organization to be successful. Executives set the company’s strategy and possess the information and experience needed to make a wide-sweeping change. Building this culture should be intentional, taking time to ask the right questions, build the team, collect and analyze the data, and to implement the uncovered insights.

Through this approach, leaders establish a framework that influences how each department operates, revising standard operating procedures that permeate every level of the organization. This differs from the often-repeated mistake of tasking the IT department with developing the data infrastructure to then send reports to various departments.

The data literate organization requires holistic changes across the company’s DNA. This visual illustrates the dimensions of management and how the data culture should be at its core:

An example that embodies the benefits of questioning your priors, rapid prototyping, and making data driven decisions comes from the book “What’s Your Digital Business Model?Opens a new window “. The authors, MIT scholars Weill and Woerner, highlight the bank BBVA’s data culture. BBVA took its customer insights and directly strategized a plan to adapt to their future vision of the industry. This included offering digitized products and improved customer service based on data. Their use of data also helped them understand their customers’ desires for new products on which BBVA capitalized.

A data literate company isn’t synonymous with being housed internally. The costs of employing the collective talent necessary can often be prohibitive, and isn’t always the best route. This goes back to knowing what you’re trying to accomplish. External experts bring extensive knowledge and skills in their area that should match if not exceed a comparable employee’s.

Experts in data and managerial integration are in high demand and can often be retained longer in an outsourced role than if hired directly. As with anything, this depends on your needs, capabilities, structure and the problem you’re trying to solve.

Start with the Right Questions

Successful data projects start with the right questions that define the target.

“You can’t hit a target you cannot see, and you cannot see a target you do not have.” -Zig Zigler

Leadership must determine where problems or opportunities exist that data could help. If you start by asking the right questions and then designing each component around the resulting goal, your project will be off to a great start. But the opposite is true if you start without these key pieces.

Here are five must-do tasks necessary to get your project off on the right foot:

1. Define the key business question(s) and resulting project goal.
2. Appoint a project lead. This person should be involved in strategic decisions so they can convey the goals throughout the organization. They must translate the significance to each department and determine their needs, plus being able to translate the technical elements of the analysis into actionable insights for strategic decision-making.
3. Audit current resources and capabilities (team, data, technology, etc.) to define what’s needed and which potential hurdles/barriers could prevent the initiative’s success. Don’t introduce additional data until the existing data is understood to avoid unnecessary complexity or confusion.
4. Determine additional resources needed to accomplish the goal (missing data, internal employees vs. outsourcing, technological gaps/inefficiencies, budget, infrastructural needs, etc.). Consultants should assist in defining all components are considered and aligned to ensure success.
5. Build the team to accomplish goals. This should include internal and external resources, ensuring that all capabilities needed are represented.

Concerning the team, it’s important not to misrepresent “data skills” as transferrable. Someone may be great at managing data and blending new data sources into a database, but this is a far different skillset than conducting advanced analysis using the right combination of techniques and methodologies to best answer the questions at hand.

Further, without expertise in extracting insights from the analysis into actionable information, the project could have an underwhelming impact. These insights need to be translated effectively so that the executive team can incorporate them into their decision-making criteria.

Most data projects fail because these tasks aren’t completed up-front. Skipping or overlooking any of these is risky, and potentially costly, both in terms of the project as well as future efforts. Building data projects from an unstable foundation can destine future efforts to fail. However, diligently ensuring that each is done correctly not only lays the groundwork for a successful project, but also a data culture that will have long-term benefits across your organization.