How To Properly Begin An Analytics Project

Topics: Analytics

How to properly begin an analytics project

Web analytics strategies and data have become an integral part of business marketing, but many organizations still struggle with setting themselves up for successful data mining within marketing projects. Even top analytics tools, such as Adobe Analytics, have limited out-of-the-box functionality, meaning that if your company is inexperienced in beginning an analytics project, the likelihood of inaccurate data and frustrated stakeholders can be pretty high. So what are the keys to preventing disaster and properly beginning an analytics project?

1. Stakeholder Approval:

It’s important to seek approval from all departments the project might affect. This likely includes marketing teams, operations, technical teams, etc. These people should sign off on the overall goal and requirements for the project, especially where it involves their work. This will help ensure gaps aren’t missed and the needs of all those affected are met. Too often we’ve seen companies running aggressive marketing projects and neglect to get complete stakeholder approval. This usually results in a lot of work, which is then for not because the work didn’t meet all the needs it should have, or connect with an approved project budget.

2. Software Inventory:

You should have a good idea of the technologies implemented across your site, how they’re integrated, and why you’re using them. Check with your technical team to ensure that your analytics software is compatible with all of these tools and that your analytics project is attainable with the systems already in place (such as how many tracking pixels are being used on your site, which tracking codes are logging visitor data, and what aspects of that data they are programmed to capture).

3. Audit the Software:

Take time to audit the pieces you identified during the software inventory phase, to ensure all parts function as they should. This will save you headaches down the road by avoiding data issues caused by unnoticed malfunctioning tools. Skipping this step could change the effectiveness of your project. Ask questions like, are all of these tools actually working the way that they are intended to or appear to be? Are you capturing the correct data from these tools accurately? Errors that you might uncover could have been caused by code installed that trumped previous code or affected other code’s data capturing abilities.

4. Establish a Data Governance Model:

You might start establishing a model with the help of those you sought stakeholder approval from. A sound data governance model includes a governing body or council, a defined set of assigned procedures, and a timeline or plan to follow by which these procedures will be executed. This ensures the project’s security, integrity, it’s participants availability, and its overall success. Clearly defining a governance model helps to keep a project on course and complete its goals in a timely, efficient manner. We’ve found that nine times out of ten, if a project is failing, it is because this step was not properly followed. Some other aspects of this model should include:

a. Defined user access and overall administration of the project. There should only be a certain set of people allowed to modify and touch things. Set your permissions appropriately so that the people who should read only, do only that. And those you actually need to write and edit can do so.

b. Drafted sprint cycles to outline points of evaluation, prioritization, logging issues, tracking requests, project history records, etc. This is no different than any other code deployment process. Have a plan for what gets rolled out when, with proper QA processes.

c. A security plan to coordinate and implement steps to ensure network security for protected data, software, and hardware.

5. Assemble the Team:

Define your data governance team based on who is responsible for future code updates, data integrations, and UI configurations to ensure all of the i’s are dotted and t’s are crossed throughout the project’s lifecycle. Team players might include:

a. Project Manager/Project Administrator

b. Web Analytics Strategist/Marketer

c. Technical Lead/Technical Strategist

d. End User (i.e. the internal resource that is running the reports and giving information to the stakeholders, rather than the outside party that might be visiting your site.)

After these steps are completed, you are ready to map out a tracking strategy to best gather data for measuring the overall success of your project. For more information on effective business web analytics strategies, email us at info@axis41.com.