Want to Improve Your Marketing Problem Solving? Think Like an Engineer.

By now marketers know they need to use AI to engage with customers in real time. However using AI is not only marketers’ biggest focus — it’s also their biggest headache. This is because marketers are great creative thinkers and strategic problem solvers, but often lack the technical skills, like coding, that traditional AI requires. To improve marketing problem solving, marketers must learn to think like engineers when it comes to AI, formulating precise prompts that make it easier to bring our creative visions and solutions to life.

Marketers often approach problems through an intuitive lens. Engineers, on the other hand, use an analytical, structured approach to make things work. When it comes to AI, engineers have been coding for years. This experience helps them understand how to effectively communicate with AI, including how to structure prompts and refine instructions to achieve desired outcomes. Recognizing the benefits of the different mindsets at play and blending the elements can help marketers think about AI, data, and technical challenges in a new way. 

The good news is there’s a new wave of AI that will make marketing problem solving easier.  Agentforce is a is a team of autonomous agents that work side-by-side with your employees to extend your workforce and serve your customers 24/7. These intelligent agents can include anything from answering simple questions to resolving complex issues — even multi-tasking. Agents turn AI from a reactive tool into a proactive assistant that takes the tedious work out of everyday marketing tasks. 

A large language model (LLM) alone relies on vast amounts of data to generate standard responses. Agents, on the other hand, require specific knowledge to address business-specific challenges. Agentforce works by giving teams tools, services, and agents that can tap into the power of LLMs and their connected business data to identify what work needs to be done, build a plan to complete the work, and then execute the plan, autonomously. Let’s take a deeper look. 

What you’ll learn

• Detailed agentic prompts, smarter marketing problem solving
• Creative problem-solving
• Visualizing
• Improving and experimentation
• Systems thinking

Detailed agentic prompts, smarter marketing problem solving

Just as a skilled engineer provides precise instructions to achieve the desired outcome, the quality of agent outputs depends heavily on the quality of input. Agents require well-structured information and clear guidance. Instead of simply providing vast amounts of data, marketers should focus on providing relevant information and clear instructions. This involves transforming human-readable information into a format suitable for the agent, ensuring the agent has the necessary context to effectively execute tasks.

The key components of an effective prompt can include:

• Clear and concise objective: Clearly articulate the goal of the task. For example, instead of “Analyze the data,” specify “Analyze the data to identify the top 5 most profitable customer segments.”
• Contextual information: Provide relevant background information, such as data sources, constraints, and any specific requirements. For example, “Analyze customer purchase history from the past year, considering customer demographics and purchase frequency.”
• Desired output format: Specify the desired output format, such as a summary report, a list of recommendations, a data visualization, or a specific action plan.
• Constraints and limitations: Define any constraints or limitations, such as budget restrictions, time constraints, or ethical considerations. For example, “Ensure all recommendations comply with data privacy regulations.”

Let’s use elements of the “Engineering Habits of Mind,” a taxonomy developed by the UK’s Royal Academy of Engineering, to break down how marketers can better prompt agents to solve everyday marketing problems, and think about AI like an engineer. (Back to the top)

1. Creative problem solving

Engineers are trained to clearly define the problem before they solve it. Rather than jumping straight into problem-solving, marketers should first ask, “What is the core challenge we’re trying to address?”

With this question at the center, marketers can think like engineers by simplifying problems into smaller, manageable parts. This might include breaking down customer journeys into stages or jobs to be done, identifying pain points in each stage, and tackling them individually.

Engineering is often described as a team sport, meaning engineers often collaborate with other engineers to solve problems. Marketers can do the same by discussing problems and creative solutions with a cross-functional team of stakeholders. 

Solve with agents: Marketers can use Agentforce as a collaborator when solving problems by asking the agent questions about the underlying data. Marketers don’t need to be able to code or write queries to explore all of the rich data they have access to. Agents use plain, natural language to ask and answer questions. Marketers can start by asking an agent to describe the data, to summarize what’s missing, and to assess any trends in the data. Then, they can begin to explore more by asking follow up questions:

1. What are the poorest performing campaigns for the last quarter? For each campaign, provide specific reasons for its underperformance.
2. What segments saw the largest attrition in the last six months? For each segment, identify potential contributing factors to churn and quantify the impact of churn on revenue for each segment.

These prompts can ultimately help marketers understand what problems to solve. (Back to the top)

2. Visualizing

Engineers are skilled at taking an abstract concept and bringing it to life. Many marketers may be adept at this too. They can dream of a campaign with a catchy email promotion, a witty tagline, or an engaging social media post and translate that vision into a concrete plan with measurable objectives. 

However, bringing these visions to life often involves numerous manual tasks and time-consuming processes. Agents simplify this process, helping marketers to more effectively translate their creative visions into successful realities with the right prompts.

Solve with agents: Agentforce helps by making campaigns incredibly easy to create, turning a vision into reality quickly and efficiently. Once marketers understand what problems to solve, they can access the agents ready-to-use skills to build an end-to-end campaign.

Using a AI prompts, marketers can generate campaign briefs, target audence segments, personalized content, and customer journeys based on user-defined goals and guidelines. Here’s how:

1. Campaign Creation: By using natural language prompts to describe campaign goals, the agent will ground that prompt in data from Data Cloud, as well as the company’s brand guidelines to generate a brief, target audience segment, email and SMS content, build a customer journey in Flow, and even provide a campaign summary. 

Let’s consider a marketer looking to build a comprehensive campaign plan for the launch of a new product. The marketer could use the prompt below to communicate to the agent exactly what they are looking for:

“Generate a comprehensive marketing campaign for the launch of our new summer product line, including target audience segmentation, creative briefs for social media, email, and display ads, and a proposed customer journey map. Consider our brand guidelines and budget constraints.”

2. Personalization Decisioning: Agentforce can help marketers scale 1:1 personalization by autonomously delivering the right content, products, and offers for each customer based on their profile.

Based on the detailed campaign the agent generated, the marketer then wants to target the identified target audience with a personalized social media messaging. This prompt to the agent could look like:

“Create five unique social media post ideas for our target audience of millennial parents, each with personalized messaging and relevant visuals. Ensure the content aligns with our brand voice and leverages current trends in parenting and family lifestyle.” (Back to the top)

3. Improving and experimentation

Engineers might be known as people who make things work, but they’d say they’re people who make things work better. An engineer’s work is never done; they’re constantly tinkering to build a better mousetrap. 

Marketers can take this to heart by continuously improving their work. Continuous improvement can take the form of testing campaigns using A/B content tests, trying out new webinar topics, delivering ads on a new social network, or cohort analysis on holdout and control groups. Best of all, marketers can use the data-driven insights to prove what works. 

Engineers often release an MVP, or minimally viable version of the product, to test the waters. Marketers can do this too, by launching smaller pilot campaigns before full-scale rollouts. This can save time and resources if things don’t go as planned. Teams can then learn from the outcomes and rapidly improve rather than waiting to launch a “perfect” campaign.

Solve with agents: One of the biggest benefits of assistive agents is freeing up a marketer’s  time by automating repetitive tasks that might involve a lot of manual work. Agents streamline workflows, making it easier to use existing marketing tools and adapt to new ones. For example, Agentforce can help marketers test their campaigns with an agent’s performance optimization skill. Post-launch, the agent can optimize paid media by autonomously identifying and pausing low-performing ads, recommending optimizations, and adjusting metrics with auto-created goals. Marketers can even write a prompt to ask the agent: 

“Based on all of the campaign data for my company, what is the best way to create an A/B test to improve engagement?” 

Agents can look at all of the customer and campaign data in the data lake, and devise an A/B test to test the hypothesis. Best of all, agents are not just for analyzing, but they can take action to execute the A/B tests quickly and efficiently. (Back to the top)

4. Systems thinking

Engineers see projects as interconnected systems. A big challenge marketers face is understanding just how everything is connected. Customers want a smooth experience, no matter how they choose to connect with a business. In fact, customers’ number one frustration is a disconnected experience. Yet, only 31% of marketers are fully satisfied with their ability to unify customer data sources.

Agentforce helps marketers overcome the fragmentation that often exists between departments and systems. Breaking down data silos is arguably the most important aspect of agents, a capability that previous AI tools couldn’t address. Agents serve as a central point of integration, pulling in customer data from various sources and ensuring a consistent brand experience across all touchpoints. Their skillsets allow customers to interact with a brand as one entity, rather than siloed departments lacking holistic data. 

Marketers can think like engineers by recognizing the interdependencies in the entire customer lifecycle, like how the whitepaper on the website can influence the sales funnel and the marketing onboarding campaigns. 

Solve with agents: Agentforce can help by following a lead from the initial lead capture through to intent, purchase, and post-sales success. Marketers can ask agents questions to analyze and summarize the lifecycles of all of their customers, to see where channels might drop off or leads fall out of the funnel. These questions could include:

1. What is the typical customer journey across all touchpoints? 
2. What are the bottlenecks or areas where customer engagement significantly declines?
3. Visualize this data and provide insights into potential causes for these drop-offs.

These insights can be used to test or adapt new campaigns.

For example, an agent can autonomously greet visitors and offer to assist them with product or service recommendations, or suggest relevant resources to learn more. This process could include capturing the contact information needed to make the recommendations more tailored, registering the customer for a webinar, providing a relevant gated asset, or scheduling a follow up appointment with a sales rep.

From the information collected, agents can then power intelligent lead nurture journeys by mapping contact information, evaluating leads based on predefined scoring models, routing qualified leads to the appropriate systems and individuals, and orchestrating tailored follow-up sequences. They can also organize flow-driven personalized experiences based on a lead’s behavior and past interactions with the brand. (Back to the top)

The mindset shift to thinking like an engineer for marketers can refresh your approach to using AI and Agentforce. Once you reframe your approach, you find yourself back at the beginning, recognizing that marketing problem solving will always be important. As a marketer, you can improve your agentic prompts and practice interacting with an agent using consumer tools. You’ll soon start to recognize the difference between effective and poor prompts and how you can improve them with more specifics. Your work as a marketer is never done, but with Agentforce, it only gets easier.