What is the greatest challenge you face when budgeting for data analytics?
Sort by:
Director of Data & Analyticsa year ago
From my perspective, D&A programs primarily encounter the challenge of substantiating their benefits. The most effective approach to navigate this challenge is to adopt a transparent and proactive stance. To ensure the successful acceptance of a proposal, preliminary work is necessary in the following areas: cost avoidance or savings, comparison of manual versus automated processes, and differentiation between intangible and tangible benefits.
Understanding the lifecycle cost of the analytics operating model is a big challenge if starting with a clean slate or if changing core data source (eg upgrade enterprise system). Leaders want the 'optimal' cost point but there are many things to consider, for example;
(1) Initially aligning on analytics objectives and determining functional requirements in terms of mandatory and nice-to-have options from both an analyst-user and information end-user perspective
(2) selecting the tech platform and data warehouse solution which means understanding fit with functional requirements, fit with current capability, license cost models, internal versus external support resources to implement and then apply and maintain the technology, annual licensing costs, etc
(3) determining internal resourcing required to meet those analytics objectives with the shortlisted tech options
(4) understanding implementation costs includes the effort from multiple stakeholders for data cleansing, mapping and migration to connect new data sources
(5) understanding security and access permissions and other enduring costs for system administration
(6) estimating effort and cost for training for data analysts, data users and administrators
Once you have a high level year-on-year understanding of the costs of your data analytics model (or model options) then you have your budget projection and can weigh those against the benefits of having - or risks of not-having - your preferred data analytics model. Quantifying those is a separate challenge!