PARTNER CONTRIBUTION
Preparing Marketing Procurement for AI
Artificial intelligence will rapidly accelerate the role of marketing procurement — here's how
In the rapidly evolving landscape of marketing procurement, data analytics and artificial intelligence (AI) are becoming indispensable tools.
The first step to leveraging AI in agency management and marketing procurement is understanding what AI can help accomplish and expedite — and what it cannot. Shrouded in the alphabet soup of artificial intelligence are powerful tools that enable entirely new ways of engaging with data and making the strategic decisions that are necessary for managing agency partnerships.
Principally, one should think of AI as the greatest marketing procurement and agency management intern in history: skilled in agency operations, project management, data processing, action planning, and, best of all, never on vacation or overloaded with tasks. While this greatest marketing procurement intern still requires oversight from a senior manager before actions should take place, AI stands to drastically reduce the effort needed in routine processing, freeing team members' time for more strategic and complex agency management activities.
Where AI particularly excels for marketing procurement is in the realm of data analysis and action planning. Here are fundamental steps to ensure procurement teams can get the most out of their data when deploying AI.
Identifying Critical Data
To make informed decisions, procurement teams need to identify the right information to use in their data analysis. This involves understanding the end-point reporting needs, mapping these to existing raw data points, identifying missing data that has to be collected, and reviewing the necessary information to support decisions.
When identifying the right information, it is crucial for marketing procurement to focus on areas such as the scope of work, agency evaluation, and production budget. These areas provide a solid foundation for understanding budget allocations, agency performance, and production costs.
But not all datasets are created equal. Often, data needs to undergo a process of review before it is ready to be trusted when making decisions against what it is saying. This process has come to be known as data preparation and cleaning.
Data preparation is a critical step to ensure that data is usable and accurate. This involves standardizing data from the start, preventing missing values, and managing outliers. There are three key steps in data preparation for marketing procurement professionals to consider:
- Standardize from the start. Avoid inconsistencies by setting standards for data collection and entry
- Avoid the blanks. Implement measures to prevent missing values
- Spot and handle the unusual. Identify and manage outliers to maintain data integrity
Regular checks and educating the team involved in data collection are essential to maintaining data quality. This includes ensuring that everyone understands the importance of data quality and how to achieve it, from agency partners to senior marketing leadership.
Analyzing and Visualizing Data
Once data is gathered, marketing teams need to apply the right category of analytics for decision-makers to use. Data analytics can be categorized into three main types: descriptive, predictive, and prescriptive. Each type serves a distinct purpose and provides unique insights:
- Descriptive analytics. This summarizes historical data to provide insights into past marketing procurement activities. Examples include analyses of scopes of work, which can help to understand budget allocations and actual spend, and agency evaluations, which aggregate and visualize ratings to pinpoint overall trends in agency performance.
- Predictive analytics. This uses historical data to forecast future outcomes. For instance, predictions based on scopes of work can forecast future budget needs based on past spend data, while agency evaluations can help in developing models to project future agency ratings based on historical performance data.
- Prescriptive analytics. This goes a step further than predictive analytics by recommending actions based on data analysis. This can include optimizing budget distribution across scopes of work and suggesting adjustments in staffing and hours to meet brand goals efficiently.
Crucial to each of these analytics categories is data visualization.
Visualization is a powerful tool to make complex data more understandable and actionable. Effective data visualization allows marketing procurement teams to tell visually compelling stories about their data that offer meaningful comparisons, reveal trends, uncover patterns, and identify outliers.
When using data visualization tools — whether basic ones such as Microsoft Excel or PowerPoint, or more sophisticated applications such as Tableau or Power BI — there are several best practices to follow:
- Design charts and tables with clarity and simplicity in mind. Visualizations should be easy to understand and avoid clutter and overcomplication.
- Choose the right visualization. It's best to select visualization types based on the data and the story being told.
- Follow basic principles of color and design. Reports should use color, contrast, and design elements to ensure readability and proper interpretation.
As AI accelerates the process for mining insights and creating data visualizations (particularly for descriptive and predictive analytics), following basic principles of good design is key to ensuring data points are interpreted correctly.
The Future of AI Is Now
As helpful as AI can be at drawing insights from data at the descriptive and predictive levels of analytics, where AI truly begins to flex its purpose and expertise for marketing procurement is in prescriptive analytics — in the collection, processing, analysis, and recommendation generation workflow.
RELATED CONTENT
Prescriptive analytics is where procurement professionals will see their next bump in productivity and effectiveness by making use of tools and advanced processing to get to recommendations and action quicker and with more accuracy. It is the next natural step for organizations to take when looking to elevate their agency management process and approach.
Through natural language processing (NLP) and retrieval augmented generation (RAG), AI now stands ready to assist even the most advanced marketing procurement and agency management teams with analyzing, reporting, and making recommendations on how best to improve the performance of their agency relationships.
But, regardless of how powerful or quick AI is in helping manage strategic agency partnerships, the process still requires a warm, human hand.
There's No Replacing the Human Touch
It is important to recognize that AI, while powerful, is not a replacement for skilled human professionals.
AI excels at automating routine tasks and processing large datasets quickly, but it still requires human oversight to ensure its outputs are accurate and contextually relevant. The role of experienced professionals remains vital when interpreting AI-driven insights and integrating them into broader strategic frameworks. There should always be a human in the loop.
Despite AI's limitations, it offers unprecedented opportunities to enhance marketers' engagement with data. By leveraging AI, marketers can expedite decision-making processes and gain deeper insights into their agency partnerships.
The power of AI lies in its ability to complement human expertise, allowing for more efficient and impactful marketing procurement strategies. The future of the marketing procurement field will be defined by this collaboration, where AI amplifies human capabilities and leads to more informed and effective decisions.
AN EYE ON AI
The Alphabet Soup of AI
Along with related fields, artificial intelligence (AI) — systems capable of performing tasks that previously required human intelligence, such as recognizing speech and interpreting data — is transforming data analytics.
Understanding key terms associated with AI is a foundational step to leveraging the technology:
- GPT (generative pre-trained transformer). AI that uses learning to produce human-like text
- ML (machine learning). Training algorithms to learn from data and make predictions or decisions without explicit programming
- NLP (natural language processing). Enabling computers to understand, interpret, and generate human language
- RAG (retrieval-augmented generation). Combines language processing and generation with data retrieval to provide accurate, relevant responses
RAG is where the key value sits for marketing procurement and agency management professionals, as it uses the power of AI, ML, and NLP to mine internal data sets including scopes of work, agency evaluation data, production data, and budgets.
By connecting an organization's agency management data to AI, procurement teams can conduct data reviews, collate datasets across years, or analyze trend patterns of spend from scope to scope, season to season, or brand to brand — completing tasks in potentially minutes that would previously take weeks.
— R.B.
Decideware is a partner in the ANA Thought Leadership Program.
