Marketing for SMEs: Keith Perhac on the Future of Marketing Measurement

I spoke with Keith Perhac, founder of SegMetrics, about what marketing tracking can realistically tell us in 2026.

We got into why leads and trials can make an advertising campaign look healthier than it really is, where UTMs still work, what happens when a customer journey lasts longer than an ad platform’s attribution window, and how AI changes the way marketers interrogate their data.

• Watch on YouTube
• Listen on your preferred podcast app

Introduction

Jeremy: Welcome back to another episode of Marketing for SMEs. Today I’m with Keith Perhac, founder of SegMetrics.

Jeremy: Keith, I’ve used SegMetrics through client accounts, and the thing that has always impressed me is the customer support. You get marketers on the call who can help troubleshoot the setup and question how the funnel itself is working.

Keith: That was a specific decision we made. Everyone on the team is a marketer, so you’re never talking to someone who is simply reading from a script.

Keith: Our main users are marketers, creators, influencers, coaches and course creators. They are often strong marketers, but they do not want the technical side to get in the way of what they are trying to achieve.

How Conversion Rate Optimisation Led to SegMetrics

Jeremy: Before SegMetrics, you ran an agency focused on conversion rate optimisation. What did CRO look like then?

Keith: The fundamentals have not changed that much. The goal is still to lead people through a customer journey that makes them feel understood and shows how your product can solve their problem.

Keith: We worked on landing pages, opt-in forms, emails, offers and the communication happening throughout the funnel. We also used quizzes and surveys to identify who was arriving and what their pain points were.

Jeremy: That wider journey matters because an ad cannot fix everything after the click. Digital Goliath’s approach to social media ads management also includes conversion tracking, landing-page considerations and reporting rather than treating the ad as an isolated object.

Keith: Understanding funnel performance is one of the most important things you can do. We built SegMetrics after dealing with millions of rows in Excel and needing a better way to understand what was happening across the journey.

Why Trials and Leads Can Give the Wrong Impression

Jeremy: The traditional media buyer who only clicks buttons inside an ad platform is disappearing. Someone managing ads now needs to understand the creative, the landing page, the numbers and what happens further down the funnel.

Keith: Anyone managing ads or anything at the top of the funnel should be looking at the entire funnel. The difference between a good and bad ad manager often comes down to whether they are willing to look past the initial acquisition and consider lifetime value.

Keith: We worked with agencies that celebrated the number of SegMetrics trials they generated. Our question was how many of those trials became paying customers and how much revenue they produced.

Keith: SegMetrics has a 14-day trial, but we look at whether someone reaches 45 days, which is effectively the second payment. That helps us distinguish a trial or an initial payment from a customer who is more likely to stay.

Keith: We also segment customers by characteristics such as industry. If one industry produces more valuable customers, that information can be used in Facebook, Instagram, Google and YouTube advertising to find more people like them.

Jeremy: This is also why our Google Ads management reporting focuses on what is creating meaningful business outcomes, not whichever platform number happens to look best.

UTMs Are Reliable, but They Cannot See Everything

Jeremy: Sometimes attribution software can look like a polished wrapper around UTMs. What is happening beyond the UTM?

Keith: UTMs are still some of the best tracking you can have because you fully control them. Nobody can change that value or take it away from you.

Keith: The limitation is that you cannot use them everywhere. They work well on ads and partner pages, but they cannot tell you that someone saw a post, watched a video without clicking or found you through an untagged Google search.

Keith: Privacy changes also shortened what Google and Meta can observe. If someone begins a trial and converts 13 days later, a seven-day attribution window may miss the sale.

Keith: SegMetrics tracks activity on the business’s own site and connects it with CRM and payment data. That makes it possible to follow a longer journey without tracking the person across unrelated websites.

Awareness and Zero-Click Marketing Remain Difficult to Measure

Jeremy: What happens when someone watches a YouTube video or sees a LinkedIn post but does not click anything?

Keith: That remains hard. A growing amount of research now happens on LinkedIn, Reddit, Instagram, YouTube and other platforms before someone ever reaches your website.

Keith: The practical options are to make the content enticing enough for the person to click through, or to connect an on-platform action with the CRM where the platform and available tools allow it. Until the person takes an identifiable action, there is limited visibility.

Keith: We see this when someone books a SegMetrics demo and remembers discovering us a week earlier. The data may show that the person first arrived from an ad four months earlier.

Keith: That gives us both the recent action that led to the demo and the earlier activity that first created awareness.

Connecting Offline Activity and Multiple Devices

Jeremy: Tracking often breaks when a customer goes offline. A person may book an appointment online, visit a gym and pay through another system later.

Keith: Phone tracking can connect an ad or website visit to a call. Another method is to create a reason for the customer to return to an identifiable digital step through an email, QR code, membership process or offer.

Keith: The visitor may initially be anonymous, but when they later identify themselves, the earlier journey can be connected with that person. The same principle helps connect desktop and mobile activity when someone clicks an email on another device.

Jeremy: It requires some orchestration rather than dropping tracking software on top of the existing process and expecting everything to work.

Product Orbit and Customer Lifetime Value

Jeremy: You have also spoken about the “product orbit.” What does that mean for an SME?

Keith: A product orbit gives the customer a logical next product. The first purchase helps them progress, which creates the next problem or need that another relevant product can solve.

Keith: SegMetrics can combine customer information with purchase data to show what different segments buy first and what they tend to buy next. If a strong pattern exists, the likely next product can be made more visible as an upsell or follow-up offer.

Using AI to Question Marketing Data

Jeremy: SegMetrics is working with a Model Context Protocol (MCP) connection so users can ask questions in natural language. Is the idea that someone can ask why a funnel dropped last week or which audience is converting?

Keith: Yes, but there is an important distinction. AI is not particularly good at receiving a raw spreadsheet without context and reliably analysing it.

Keith: The SegMetrics MCP server runs an actual report query in the background and then gives the structured result to the AI for interpretation. The platform still performs the processing and number crunching; AI helps the user understand the result and investigate further.

Keith: Someone could ask which industries produced the strongest customers, identify an underperforming segment and then go several levels deeper into the behaviour behind it. We have also trained the system to push back on weak questions and help people become better analysts.

Why Vibe-Coded Tracking Is Not Necessarily Free

Jeremy: Does vibe coding threaten established software when someone can build the 10% they personally need?

Keith: In the short term, yes. I use the phrase “free as in puppies”: getting started may be free, but then you have to maintain it, fix it and support it for years.

Keith: A simple tool might show where someone came from or whether they entered ActiveCampaign. As more tracking requirements are added, one change can break an earlier feature and the owner becomes responsible for maintaining the entire system.

Keith: We still support SegMetrics features introduced 13 years ago because customers rely on them. That ongoing reliability is very different from building a simple demonstration that works well enough for a social media post.

Key Takeaways

• A lead or trial is not proof that an advertising campaign is profitable. Measurement needs to follow the customer through revenue, retention and lifetime value.
• UTMs remain reliable because the business controls them, but they cannot capture every zero-click interaction, untagged source or offline step.
• Offline activity becomes easier to attribute when the customer returns through an identifiable action such as a phone call, email click, QR code or membership process.
• AI is more useful when interpreting structured, verified reports than when it is asked to make sense of raw data without business context.
• A quickly built tracking tool may solve a narrow problem, but maintaining it as requirements grow is a separate cost and responsibility.

Learn more at SegMetrics and connect with Keith Perhac on LinkedIn.

This conversation was recorded for Marketing for SMEs with Jeremy Yang. You can watch the full episode on YouTube or listen on your preferred podcast app.