In this episode of the Jukebox Podcast from WP Tavern, Nathan Wrigley speaks with Matt Schwartz, founder of Inspry, an Atlanta WordPress and WooCommerce agency, and creator of CheckView, a WordPress testing product. Matt has run an agency since 2011 and has recently focused much of his attention on practical uses of AI inside agency operations — not just for generating sites or content, but for improving workflows, processes, and the value agencies deliver.
Why focus on process, not just output
A key theme Matt stresses is that AI’s biggest impact for agencies will be on process, not merely on building pages. Execution—creating a brochure site or producing content— is increasingly becoming commodified. As AI tools make routine tasks faster and cheaper, the differentiator for agencies will be how they run projects, communicate with clients, and create measurable value. That means agencies should invest energy in refining SOPs, client touchpoints, and internal operations, and then consider where AI can boost those areas.
Transparency and ethics
Matt’s stance on disclosure is pragmatic: if AI is used in a client’s product, tell them. When AI is used internally to power processes that don’t directly affect client-facing deliverables, it’s not always necessary to flag it. The ethical line he draws is about client outcomes and expectations — clients should know if something on their site or in their deliverable was generated by AI, and they should never be surprised by poor or inappropriate output.
Why now?
The pace of change has accelerated: what used to be incremental improvements across tools is now frequently seismic. Clients are aware of AI capabilities and expect faster turnaround and lower cost for routine work. That raises the floor: agencies that don’t look at AI risk losing competitiveness on speed and process. But Matt cautions against blindly adopting every new tool. The right approach is intentional: evaluate where AI aligns with your agency’s values, capabilities, and client needs.
An AI vision document
Matt recommends creating an AI vision document for your agency — a written guide that outlines where AI will and won’t be used, sets guardrails, and defines human review points. The document should map your key processes (sales, proposals, PM, QA, support, launch) and identify repetitive or data-heavy tasks where AI could save time. It should also flag tasks that require human judgment and should remain human-only.
Practical uses and examples
– Spotting patterns: Use AI to analyze time-tracking, tickets, and other operational data to find repetitive tasks worth automating. AI is strong at pattern recognition and can point teams to low-hanging improvements.
– Meeting summaries and proposals: AI can transcribe meetings, extract action items, and draft tailored proposals or SOWs by combining meeting notes, a client’s website, and your prior proposals. This enables higher-quality, client-specific proposals without spending excessive time on each draft.
– Support workflows: Instead of replacing support staff, AI can augment them. By ingesting ticket history, error logs, and site data, automation can prepare context-rich summaries and likely resolutions for human review, speeding up response time and triage.
– Debugging and site management: Connecting AI agents to a site’s APIs or logs can dramatically reduce time spent diagnosing one-off problems. AI can propose likely causes and fixes, letting humans validate and finalize changes. Page builders and WordPress core improvements are making it easier for agents to understand structured content and perform edits — but Matt emphasizes human review before publishing.
Productizing AI expertise
If your agency is already building AI skills, those skills can be turned into a service offering. Rather than advertising “AI,” sell the outcome: process automation, custom lightweight internal apps, workflow improvements, or recurring managed automation. Clients don’t always want the word “AI” up front — they want results: save money, increase efficiency, or automate repetitive tasks.
Be careful with custom apps and security. While AI can lower the cost of building lightweight custom tools, agencies must still do due diligence on security, maintenance, and long-term support. For many clients, using an existing SaaS may still be the right choice; for others, a tailored automation or app built intelligently with human oversight can be compelling and profitable.
Risks and human-in-the-loop
Matt repeatedly stresses the importance of human oversight. AI can be fast and adept at pattern recognition, but it can also hallucinate or produce unsafe changes. For sensitive moments — customer support during a crisis, complex debugging, or final content and code edits — agencies should define review checkpoints and never fully outsource judgment to an agent.
WordPress as scaffolding
The conversation highlights how WordPress is evolving into reliable infrastructure that agents and AI can work against. With better APIs and growing support for structured editing, WordPress can remain the platform while AI-driven agents manage content edits, translations, or site tweaks. This reduces friction for non-technical stakeholders and changes how teams interact with the CMS: increasingly via AI-enhanced workflows rather than manual admin edits.
A pragmatic approach for agencies
– Audit your processes first. Look for repetitive, time-consuming tasks and ask if AI can add value there.
– Create an AI vision document to set scope and guardrails.
– Be transparent with clients when AI affects deliverables.
– Productize the outcomes (automation, internal apps, process optimization) rather than selling “AI” as a buzzword.
– Keep humans in the loop where risk or judgment is high.
This episode is the first of a two-part series. Matt and Nathan cover eight of sixteen show-note items here, ranging from strategy and ethical considerations to concrete operational changes. Part two will continue with more examples and deeper dives into how agencies can safely and effectively implement AI.
If you run or work in a WordPress agency, this conversation is a practical invitation: treat AI as a tool to optimize process and expand the kinds of services you can offer, but do so intentionally, with documented policies and human oversight.
