AI Strategies That Save Time and Money for Agency Owners

Curious how AI can make a difference in your marketing agency, beyond generating content? 

In this article, a leading AI expert shares AI tools and real-world examples you can start using immediately.

Why Should Agency Owners Embrace AI?

Artificial intelligence (AI) is transforming businesses across every industry, and marketing agencies are no exception. 

As AI capabilities like natural language processing and machine learning rapidly advance, agencies can stay competitive and leverage these technologies to work smarter, deliver more value to clients, and future-proof their operations.

As the co-founder and chief data scientist at Trust Insights, Christopher Penn offers practical marketing strategies for agencies looking to save time, money, and effort with the help of AI technology.

AI Strategies for Marketing Agencies to Save Time and Money

While AI is not a new technology—it has been around since the 1950s—recent advancements have made it much more accessible and applicable for marketers. 

Ten years ago, working with AI required specialized coding skills; now, user-friendly software-as-a-service (SaaS) tools provide access to powerful AI capabilities through simple interfaces.

As Chris explains, this increased accessibility means AI can now benefit marketing strategies across every industry vertical. Whether an agency works in healthcare, finance, retail, or beyond, AI marketing tools that process and generate language can provide value. 

The key is appropriately evaluating if a particular AI platform matches the problem your agency needs to solve.

#1: First, Ask the Right Questions

People often ask, “What AI marketing tool should I use?” But that's the wrong question. You wouldn't ask, “What blender should I use?” without first figuring out what you're trying to make.

Start with your goal first, not the technology. What problem are you trying to solve? Then, figure out the process and skills needed to achieve it. Otherwise, you risk choosing tools that don't help achieve your goals, like using a blender to cook steak.

Chris recommends the “5 Ps” framework when deciding which AI-powered tools to implement: 

Purpose: What problem are you trying to solve?

People: Does your marketing team have the proper skills?

Process: Do you have an effective process or “recipe” to achieve your purpose?

Platforms: Only after the first 3 Ps should you pick AI tools and technologies.

Performance: Did the AI solution actually solve the original problem and purpose?

This approach ensures the technology you add benefits your business.

#2: Rethink Agency Business Models in the AI Era

Many agencies still rely on billing clients by the hour, but AI technology threatens to turn this model upside down. AI now completes tasks in minutes that previously took marketing teams hours or days to complete.  

As Chris warns, “Your agency is in mortal danger from AI if you bill by the hour.”

He advises agencies to shift focus from the time spent to the value created when billing clients. This adjustment may require agencies to reevaluate internal processes and workflows that AI can augment or automate. 

The goal should be to free up human employees from repetitive tasks so they can focus on higher-value work.

Building an internal AI council is one approach Chris suggests to help agencies continually identify new applications for AI and test ideas through rapid prototyping sprints. This council would research available tools and assess opportunities for implementation based on the agency's people, processes, and clients.

#3: Evaluate AI Tools

With new AI platforms launching constantly, how can agencies evaluate options? 

Chris recommends starting close to the source by leveraging large language model (LLM) foundations like Claude, ChatGPT, Google Bard, and Microsoft Bing. Their developers frequently update these foundational models to expand capabilities over time.

Vetting platforms built on these foundations requires asking questions about their underlying language model architecture and flexibility. If the vendor platform does not readily support swapping in improved language models as they are released, their technology risks becoming outdated.

For handling sensitive client data under non-disclosure agreements, Chris points to a free, open-source tool called LM Studio, which allows you to run LLMs locally on your agency's computers without transmitting data externally. This on-premise approach satisfies strict data protection requirements for industries like healthcare.

#4: Streamline Contract Reviews and Creation

Dealing with legal agreements like NDAs and master service agreements (MSAs) requires precious time across marketing teams. AI comes to the rescue by reviewing agreements and even generating contract language.

Chris recommends using Claude, Anthropic’s AI assistant, to review incoming NDAs or MSAs. Simply upload the document to Claude and ask targeted questions to surface potential issues, like:

• Is this agreement bilateral and fair to both parties?
• Does the confidentiality clause fully protect my interests?
• Is the length of the agreement (e.g., eight years) excessively long?

Based on its analysis, Claude will flag provisions that could pose problems if you were to sign. Agencies can quickly strengthen one-sided clauses by instructing Claude to “rewrite section X to be more bilateral.”

The time savings quickly compound since AI technology reviews surfaced all significant issues in minutes, compared to hours of human review. Agencies see dramatic time and cost reductions.

For creating agreements, agencies can likewise prompt Claude or other AI-powered tools to generate neutral contract language, NDAs, privacy policies, and other documents. After the AI produces an initial draft, your legal team can rapidly finalize contracts rather than create them from scratch. 

Even though AI has perfect memory, Chris emphasizes that AI should only partially replace lawyers when finalizing official agreements. Think of AI as providing an advanced starting point to minimize manual drafting.

#5: Optimize Content Creation

Given their reliance on compelling content, only some areas offer more potential for AI optimization than copywriting and design at marketing agencies.

However, Chris notes a common misconception about AI and copyright. Under current law, content entirely created by AI algorithms cannot be copyright-protected since no human authored the work. This poses serious risks for agencies claiming copyright on client deliverables.

To retain copyright while leveraging AI efficiency, Chris recommends a human-in-the-loop approach:

1. Use AI tools like Claude for ideation and outlines
2. Humans review AI outlines for inaccuracies
3. Humans create final content with AI guidance

With this process, the AI provides a framework while humans still actively author content, allowing agencies to uphold copyright. By handling tedious upfront research and ideation, AI enables copywriters to focus their skills on high-value writing rather than starting from a blank page.

For images and multimedia, agencies should follow the same human-in-the-loop method for copyright protection. AI can suggest creative directions and concepts, while humans should design and create. 

Getting More from Customer Data 

With access to expansive customer data from sources like online reviews and social conversations, agencies often need help to extract meaningful insights. Powerful language AI saves the day, according to Chris.

First, extract all customer reviews, surveys, etc., and then feed them into an LLM. Ask the AI for a top-level summary of things, like:

• 5 top positive themes and their frequency
• 5 top negative themes and their frequency
• 3 recommended fixes 

This summary becomes an executive briefing on the customer experience. It leverages your customers’ voices to spotlight strengths, weaknesses, and areas for improvement; this process takes minutes vs hours of manual analysis. 

Similarly, using AI tools, you can leverage authentic customer voices for better marketing campaigns. 

For example, collect your best 5-star reviews across sites and channels. Paste them into a document to create training data for AI. Then, use that data to generate social media posts. 

Ask the AI to write Instagram captions, TikTok videos, etc., based on actual customer perspectives.

Even your most flexible, scalable copy can fail to connect to your target audience. Using the authentic voice of real customers performs much better. 

Streamline Analysis and Reporting

Reporting represents another prime area where AI can pare down the manual workload for agencies.

Chris suggests first exporting raw data from analytics platforms like Google Analytics to analyze client campaign data and performance. 

Next, use an AI coding tool like GitHub Copilot to write custom scripts that process the exported data. For example, ask Copilot to summarize weekly site visitors and eCommerce conversion rates by traffic channel. Then, export the summaries into easy-to-update spreadsheets.

Chris recommends snapping screenshots of key platform dashboards to build complete client reports and uploading them into Google Bard. Simply ask Bard to interpret the platform’s dashboard and describe the top takeaways. 

Then, edit Bard’s analysis for client context. By generating first drafts based on dashboard data, agencies save analysts hours of manual interpretation.

The key is knowing where AI adds value to your agency.

#6: Adhere to Ethical and Legal Standards

With great power comes great responsibility. Following ethical and legal protocols is essential as agencies deploy AI more extensively within their organizations. 

As Chris states, “You never, ever let AI output directly be handed to a client” without human review. Consider even advanced AI as a very competent intern requiring guidance and oversight before client delivery.

While legal requirements continue evolving, Chis offers specific best practices agencies should adopt when leveraging AI:

• Disclose the use of AI to clients for transparency.
• Anonymize data by removing personal identifiers. Personal identifiers are either sensitive protected information like an IP address, email, social security number, etc., or protected health information like a person’s weight, blood pressure, whether they’ve had COVID-19, etc. 
• Cleanse data of sensitive information that is not essential for the AI.
• Always have humans review output before client delivery.
• Continuously assess AI platform capabilities against client contracts.

Additionally, properly securing confidential business and customer data remains imperative. 

For example, you cannot use a third-party service like ChaptGPT if your client data contains anything under an NDA. However, you can use Meta’s Llama or any LM Studio models because they run locally on your computer.  

If contracts promise clients full ownership over intellectual property (IP), using AI could violate these agreements. Ensure the contracts you sign have confidentiality data protection and IP clauses.

Carefully assess data privacy implications before using AI. With access to customer conversations and behaviors, social media data can easily contain protected health information and other sensitive data, especially globally across regions like the European Union and China. 

Say you collect social media posts to train AI. If even one post discloses someone's vaccine reaction or health issue, the whole dataset qualifies as protected health information. Whether intended or not, you must treat that data with privacy safeguards for medical details.

Always scrub datasets of personal identifiers where possible. Agencies risk significant fines or even criminal liability for mishandling sensitive information. When in doubt, consult expert legal guidance to navigate data protection compliance safely as regulations differ across jurisdictions.

The Future of AI in Marketing

While AI enables agencies to complete discrete tasks more efficiently, Chris predicts AI capabilities will soon become ubiquitous across all software tools and platforms. For example, Microsoft is incorporating Copilot into Office and Windows, and Google Workspace recently integrated Duet.  

Just as computers evolved from single-core to multi-core processors, AI networks provide more nimble cooperative processing. Powerful devices like Apple's latest iPhone, packed with neural processing chips, will also increase access to on-device AI for professionals on the go. 

Given these trends, marketing agencies have no choice but to make AI expertise a core competency within their organizations. Leaders must retool their agency's purpose, people, processes, and performance metrics to stay fit for the AI-powered future.

As repetitive language-based tasks will become fully automated as current tools like ChatGPT evolve, agencies must be fluent in prompting AI platforms as part of their daily workflow.

Chris recommends auditing your current systems and workflows to pinpoint opportunities to hand rote responsibilities to AI. “AI is not going to take your job. A person skilled with AI will take the jobs, plural, of people who are not skilled with AI,” Chris believes. 

More importantly, agencies need a long-term vision for harnessing AI to provide value, not just save time. 

Consider how augmented capabilities can open up new revenue streams or client offerings. What future projects could you take on by reclaiming lost time from AI efficiency? Training staff on conversational prompting skills also ensures your team evolves as new AI capabilities emerge.

Most critically, agencies must evolve their business and pricing model for an AI-powered world. Your agency could lose revenue without alternative value pricing when AI effortlessly replaces hundreds of billing hours. 

Agencies must demonstrate and charge for the expertise, creativity, and strategy they provide clients above essential execution. By specializing in high-value skills only humans offer, agencies future-proof their value.

Where to subscribe: Apple Podcasts | Spotify | YouTube Music | YouTube | Amazon Music | RSS

✋🏽 If you enjoyed this episode of the Marketing Agency Show, please head over to Apple Podcasts, leave a rating, write a review, and subscribe.

---

Stay Up-to-Date: Get New Marketing Articles Delivered to You!

Don't miss out on upcoming social media marketing insights and strategies! Sign up to receive notifications when we publish new articles on Social Media Examiner. Our expertly crafted content will help you stay ahead of the curve and drive results for your business. Click the link below to sign up now and receive our annual report!