In Clear Focus: Customer Reviews as Strategic Intelligence with George Swetlitz
Host Bio
Adrian Tennant is the Chief Strategy Officer at Bigeye, a strategy-led, full-service creative agency, where he helps guide brand and marketing strategy through audience development and consumer-insight work. (Bigeye) He joined Bigeye in 2019 as Vice President, Insights, bringing experience spanning digital strategy and research leadership, including serving as Vice President/Director of Digital Strategy at 22squared and as Chief Experience Officer/Director of Research at Blue Kite Insight, an advertising and user-experience startup he co-founded. (Bigeye) In 2022, Bigeye promoted Tennant to Chief Strategy Officer, citing his leadership in expanding the agency’s audience development practice across quantitative, qualitative, and mixed-method studies that inform media planning and creative strategy. (Bigeye) Tennant also hosts Bigeye’s weekly podcast In Clear Focus, which explores trends in advertising, marketing, and consumer insights. (coverings2025.eventscribe.net)
Summary
Adrian Tennant frames customer reviews as more than a reputation score—they’re a high-signal, high-volume dataset that most organizations underuse. Early in the conversation, he sets the stakes plainly: reviews are “a flood of unstructured data that’s difficult to analyze and act upon,” and the episode asks, “What if customer reviews could become one of your most valuable sources of strategic intelligence?” (Tennant).
From there, Tennant’s questions keep returning to a practical marketing premise: reviews sit unusually close to conversion, and the businesses that treat the review ecosystem as a living customer conversation—not a box to check—gain an advantage. George Swetlitz (co-founder of RightResponse AI) reinforces that “the cheapest and most effective way to grow is to have people call you,” and says that for location-based businesses, “the center… is the review ecosystem.” Swetlitz also makes the bottom-of-funnel point that anchors much of Tennant’s interest: “There’s no one closer to the bottom of the funnel than someone actually reading your reviews,” and “More people read reviews than visit websites.”
Turning review text into “mini-ratings” leaders can act on
Tennant challenges the limitations of star ratings directly: “Traditional review platforms often provide a single rating that doesn’t tell the whole story.” In response, Swetlitz explains how RightResponse AI approaches “review sentiment analysis” (and related competitor views) by converting messy review narratives into clearer operational signals.
Instead of treating each review as one number, Swetlitz describes breaking reviews “down into phrases,” then using AI to assign those phrases—“in their context”—to customizable topics. The output is what he calls “mini ratings,” expressed as “percent positive mentions” per topic. His example is intentionally concrete: if there are 100 mentions of food quality and 90 are positive, the business sees “90% positive on food quality.” The reason he believes this matters is interpretability: “If you have a 4.8, well what’s a 4.8? But if you have a 94% positive mentions, you know exactly what that is.”
In the episode’s throughline, this topic-level view turns reviews into operational diagnosis. Swetlitz explains that leaders can compare locations, brands, or regions visually, topic by topic. He describes showing onboarding clients a chart where a location is “underperforming on every single topic,” and contrasting that with what teams used to rely on: a vague sense that one store’s “average rating… is 4.4” versus “our average is 4.7.” Tennant’s focus here is organizational: how executive teams can use this kind of review intelligence to “identify operational issues before they escalate.”
Why response quality matters: reviews as a conversion conversation
A major portion of Tennant’s interest is how review responses influence real buyers, not just algorithms. When Tennant asks how RightResponse AI generates responses that are personalized rather than generic, Swetlitz starts with an observation Tennant implicitly validates by pursuing it: customers are increasingly intolerant of automation that feels fake. Swetlitz says, “you increasingly see people being frustrated with generic AI responses,” noting that people will reply to businesses with complaints like, “you are just using a bot.”
Swetlitz’s central argument is that a review response is not merely housekeeping—it’s part of the sales conversation happening in public. He calls the marketer’s challenge “how do I bring my website to the review conversation?” and says that when a business “provide[s] people with useful and helpful information” in its reply, it can influence both the reviewer and the “review readers that are trying to decide where to go next.” In that framing, template replies “say nothing,” and generic AI that simply mirrors the review “says nothing,” because it fails to add business-specific value.
The “facts” approach: making AI responses business-specific
Swetlitz then explains how RightResponse AI tries to avoid generic “AI slop” by grounding responses in business-specific “facts.” He describes “a whole series of AI agents” that touch every incoming review. Some are basic—like deciding whether a reviewer’s name is actually a usable name. Others are more structural, like a “fact system” built by analyzing a business’s last 500 reviews to identify recurring themes and create usable, on-brand guidance.
His illustrative example is operational and service-oriented: if reviews repeatedly mention difficulty finding parking, a fact might instruct the response to tell customers that “parking for the building is behind the building off of Beach Street.” Swetlitz’s claim is that these kinds of details help a business “participate in that conversation and provide a useful response,” rather than a generic acknowledgment that reads like automation.
Personalizing review requests to improve conversion
Tennant also explores how the platform applies personalization upstream—getting more reviews, and improving the likelihood that requests convert. Swetlitz says the “first part” is “conversion of the request into a review,” and argues that “the more you can make an emotional connection,” the more likely the customer is to respond.
He contrasts generic outreach (“hi, we saw you yesterday”) with a message that includes real context: a named staff member, what was completed, even “a photo of the finished project,” plus prompts customers care about. Swetlitz says RightResponse AI “tie[s] in to the CRM systems that have information about that customer” and uses AI to personalize the request accordingly.
Case examples Tennant calls out as strong proof
When Tennant tees up practical examples—“we love case studies”—Swetlitz shares scenarios that emphasize multi-location execution, limited staffing, and measurable improvements:
- A large home services business distributed review-response responsibility to regional managers. Responses were “auto generate[d]” (not auto-published), and managers typically accept them with minor edits—Swetlitz says “85%… they just accept.” He reports performance changes from roughly “90% percent positive” to “95% positive,” alongside a jump from “about 20%” response coverage to “80% response rate,” and adds, “they’ve told us their business is booming.”
- A business serving groups (alongside individuals) used a personalized review request featuring “a photo of the group” to reconnect customers to the experience; Swetlitz says they “tripled their conversion rate,” moving from “sub 10%” to “about 30%.”
- Another organization with “more than a hundred locations” had low response quality and “sub 50%” response rate with only two people managing reviews. Swetlitz reports that with the system they reached “a hundred percent response rate,” and that “90% of the time” now goes to “real issues with real customers,” supported by “one FTE.”
Tennant’s reaction reinforces the credibility value of these examples: “Yeah. Great examples.” (Tennant).
AI skepticism vs. AI that feels more human
Late in the episode, Tennant raises the tension many operators feel: how to balance automation and authenticity. Swetlitz draws a sharp line between “fake personalization” and what he calls “real personalization.” He gives a cautionary story about receiving a sales email that referenced Penn and recommended a restaurant that “only opened maybe 25 years after I graduated.” For Swetlitz, that’s the risk: AI hunting for a conversational hook that collapses under basic scrutiny.
By contrast, he argues that real personalization is “controlled by the business,” rooted in “real things,” and helps “the people who know things” extend themselves when they “can’t be everywhere at every minute.” Tennant explicitly aligns with that distinction in a brief but clear endorsement: “Hmm. Agreed.” (Tennant).
The strategic takeaway: sophisticated review engagement wins
In closing, Swetlitz predicts review volume will keep rising and argues companies will be forced to decide whether to treat it as noise or intelligence. He says businesses that engage the review ecosystem “in a sophisticated way… will win over companies that don’t,” and reframes the work as commercial, not administrative: “this isn’t just a task. This is a conversion engine.” Tennant’s final tone matches the episode’s arc—curious, operationally focused, and affirming—ending with, “Hmm. Great conversation.” (Tennant).
Q&A
1) Are online reviews really more important than my website for winning customers?
Swetlitz argues reviews sit unusually close to the decision moment: “There’s no one closer to the bottom of the funnel than someone actually reading your reviews.” He adds, “More people read reviews than visit websites,” which is why he frames the review ecosystem as central to organic growth for location-based businesses. Tennant reinforces the strategic angle by describing reviews as “a flood of unstructured data that’s difficult to analyze and act upon.”
2) Should I respond to every review, or only the negative ones?
Swetlitz’s examples point toward responding more consistently, not selectively. He describes one business moving from responding to “about 20%” of reviews to an “80% response rate,” and another reaching a “hundred percent response rate.” Tennant signals approval of the direction with, “Yeah. Great examples.” Swetlitz also clarifies the workflow he describes is assisted—responses can be drafted automatically, “not auto publish,” with humans reviewing.
3) If I use AI to reply to reviews, how do I avoid sounding robotic?
Swetlitz says customers are increasingly annoyed by “generic AI responses,” and notes they’ll sometimes call it out directly: “you are just using a bot.” His recommended remedy is to avoid replies that simply mirror the review and instead add business-specific, useful information—because templates and generic automation “say nothing” in a real customer conversation.
4) What does an “authentic” review response mean in practice?
Swetlitz defines authenticity as replying the way a real person would: “respond to them in the way that you would if you were speaking to them face-to-face.” He connects that to participating meaningfully in the review conversation, not merely acknowledging it. Tennant’s questions keep the focus on why this matters: reviews function as a marketing and conversion touchpoint, not just a reputation task.
5) What kinds of details make a review response feel real instead of generic?
Swetlitz describes using business-specific “facts” that can be woven into responses when relevant—especially practical details that help the next customer. His example: if reviews repeatedly mention parking frustration, the response can include specific guidance (where to park). He positions that kind of specificity as “a helpful and useful piece of information,” rather than a generic thank-you.
6) I manage multiple locations—how can reviews help me spot operational problems before they get worse?
Tennant asks this explicitly for multi-location and franchise operations, focusing on early detection. Swetlitz describes comparing locations/brands/regions visually by “percent positive by topic,” which can make it obvious when one location is “underperforming on every single topic.” He contrasts that with relying on a vague rating gap (e.g., 4.4 vs 4.7), which is less diagnostic.
7) My star rating is high, but I still feel like we’re missing something—how do I diagnose what’s actually driving reviews?
Tennant calls out the limitation: a single rating “doesn’t tell the whole story.” Swetlitz explains reviews often contain mixed signals (“I love the burgers, but I hated the fries”), so he prefers breaking reviews into phrases and assigning them to topics. That yields topic-level signals that can reveal where performance is slipping even when the overall average looks fine.
8) What should I track besides star ratings if I want real “Voice of Customer” insight?
Swetlitz emphasizes “percent positive mentions” by topic as a clearer operational metric. He gives an example: if there are 100 mentions of food quality and 90 are positive, that topic is 90% positive. He argues that’s more interpretable than an abstract average: “If you have a 4.8, well what’s a 4.8?”
9) How do I turn thousands of review comments into something leaders can actually use?
Tennant frames the core challenge as unstructured feedback that’s hard to analyze. Swetlitz’s approach is to break reviews into phrases and classify them to topics “in their context,” producing structured outputs leaders can compare and act on. He also describes an “AI generated monthly review analysis” synopsis used by regional managers, intended to make review learning continuous and digestible.
10) Do review responses actually affect conversion, or are they just “nice to have”?
Swetlitz argues responses matter because review readers are already deciding: “They are deciding and they’re looking for a conversation.” He frames the marketer’s challenge as “how do I bring my website to the review conversation?” and says helpful responses can influence not only the original reviewer but also “all the review readers” evaluating where to go next.
11) How do I handle negative reviews at scale without ignoring real customer problems?
Swetlitz describes a large, multi-location business where two people couldn’t keep up; response quality and response rate were low. He says that after adopting their system, they reached “a hundred percent response rate,” and “90%” of the team’s time shifted toward “real issues with real customers,” because routine responses were easier to handle and attention could be reserved for customers who truly needed help.
12) Is it risky to auto-publish AI replies, or should someone always review them first?
Swetlitz describes an approach where responses are drafted automatically but reviewed by humans: “auto generate… not auto publish.” In his example, regional managers typically accept the drafts—“85%” of the time—otherwise making minor edits before publishing.
13) How do I get more customers to actually leave reviews when I ask?
Swetlitz says the biggest lever is improving conversion of the request itself by making an emotional connection and including real context. He contrasts generic outreach (“hi, we saw you yesterday”) with a message that includes specifics and even “a photo of the finished project.” He also shares a case where a group-oriented business used a photo of the group and tripled conversion (from sub-10% to ~30%).
14) What’s the difference between personalization that helps and personalization that backfires?
Tennant raises the tension between automation and authenticity. Swetlitz warns against “fake personalization,” illustrating it with an email that referenced a restaurant that opened decades after he graduated—proof that AI can manufacture relevance incorrectly. He contrasts that with “real personalization” grounded in real business/customer context and controlled by the business. Tennant reacts with clear alignment: “Hmm. Agreed.”
15) People are skeptical of AI—how can AI replies make a brand feel more human, not less?
Swetlitz acknowledges the backlash against generic AI and argues the goal isn’t to add synthetic friendliness; it’s to extend real knowledge and values consistently. He frames it as helping when “the people who know things can’t be everywhere at every minute,” while keeping responses grounded in accurate, business-specific facts.
16) What review metrics should I watch if I’m trying to improve operations, not just reputation?
Swetlitz points to topic-level percent-positive signals as operational indicators—e.g., seeing “service” at 95% positive but “quality” at 70% positive tells leaders what to fix. Tennant’s executive-focused questioning ties this to early detection and accountability across locations before problems “escalate into much larger problems.”
17) If I’m implementing AI in a customer-facing workflow, what makes it reliable instead of erratic?
Swetlitz shares two practical lessons: “smaller is better” (don’t overload a single prompt), and use a chain of agents rather than one do-everything prompt. He gives an example of troubleshooting their own fact-generation workflow: they realized they were asking the model to do math, and “language models don’t do math,” so they moved math to the backend.
18) Can a marketing agency use review data to do audits or prep pitches without weeks of manual research?
Swetlitz describes agencies using the tooling for research—scraping/downloading reviews for multiple locations and competitors and running sentiment analysis—so they can prepare insights quickly instead of having “a junior person spending a week” doing manual review reading. Tennant’s question frames this as a practical agency workflow (including white-label considerations).
19) Is this kind of review intelligence only for physical locations, or can it apply to other businesses too?
Tennant notes the platform is designed for businesses “with physical locations” and asks about ecommerce. Swetlitz says that extension isn’t available “right now,” but he adds many non-location businesses still rely on platforms like “Trustpilot” and the “Better Business Bureau,” and he says those are supported.
20) Are reviews just a reputation chore, or can they actually be a growth engine?
Swetlitz explicitly rejects the idea that reviews are merely a task: “This isn’t just a task. This is a conversion engine.” He argues that companies that engage the review ecosystem “in a sophisticated way… will win,” framing it as a competitive advantage. Tennant’s opening lens supports the same direction, positioning reviews as “strategic intelligence,” not administrative cleanup.
Transcript
[00:00:00] Adrian Tennant: Coming up in this episode of Incl Focus.
[00:00:03] George Swetlitz: There's no one closer to the bottom of the funnel than someone actually reading your reviews. More people read reviews than visit websites, so the challenge for a true marketer is to say, how do I bring my website to the review conversation?
[00:00:24] Adrian Tennant: You are listening to in Clear Focus, fresh Perspectives on marketing and advertising. Produced Weekly by Big Eye, a strategy led full service creative agency growing brands for clients globally. Hello, I'm your host, Adrian Tennant, big Eye's chief Strategy Officer. Thank you for joining us. Every day, millions of customers leave reviews online, but for most businesses, this feedback represents a flood of unstructured data that's difficult to analyze and act upon.
[00:00:58] Adrian Tennant: What if customer reviews could become one of your most valuable sources of strategic intelligence? Well, our guest today believes they can. George Switz is the co-founder of Right Response ai, an artificial intelligence platform that transforms how businesses understand and respond to customer feedback.
[00:01:19] Adrian Tennant: A Harvard Business School graduate with experience at McKinsey and Company, George understands both the operational challenges of multi-location businesses and the power of AI to solve real marketing problems, to discuss how customer reviews can become strategic intelligence. The challenges of reputation management at scale and why AI powered personalization is replacing generic automation.
[00:01:46] Adrian Tennant: I'm delighted that George is joining us today from Jacksonville, Florida. George, welcome to Incl Focus.
[00:01:54] George Swetlitz: It is great to be here. Thank you very much for having me.
[00:01:57] Adrian Tennant: Well, George, before we dive into Right response ai, could you tell us a bit about your career?
[00:02:04] George Swetlitz: Sure. So I've had a long career. I started off in, uh, consulting, so I worked for McKinsey and Company, very large consulting firm, and then transitioned into operating roles, and so I, I ended up leading a division.
[00:02:19] George Swetlitz: For a company called Sarah Lee Corporation, and then went back on the consulting side and spent a number of years working with primarily large organizations to improve overall profitability. Then I started working more closely with private equity firms and ended up becoming the CEO of a private equity backed consolidation.
[00:02:42] George Swetlitz: In the audiology industry, and that was alpaca audiology, which was right before I founded Wright Response ai.
[00:02:51] Adrian Tennant: Got it. George, what led you to start Wright response ai,
[00:02:55] George Swetlitz: so as CEO of a 220 location business. One of the key things that you focus on is growing organically, so that to me is having people call you as opposed to you reaching out through paid ads and in other ways.
[00:03:17] George Swetlitz: The cheapest and most effective way to grow is to have people call you. And what we learned was at the center of that is the review space for businesses that are location based. The center, the core of getting people to call you is the review ecosystem. That's the core. And so we focused on how to do that.
[00:03:38] George Swetlitz: We spent a lot of time thinking about it and working with vendors to do that in a very effective way. And the problem was we really couldn't find good vendors to help us. And of course, this is before ai. They were very expensive, and the outputs that you would get from these systems were not very good.
[00:03:56] George Swetlitz: We ended up selling the business in 2021, and then in 2022, jet G PT came out and I thought this might be a good way to solve the problems that we had. And so I got the team together and we started right response. We spent about a year, year and a half building the platform and launched in late 23.
[00:04:18] Adrian Tennant: Can you explain how right response AI transforms customer reviews into strategic intelligence?
[00:04:27] George Swetlitz: Yes. And this is particularly important for multi-location businesses because what you're trying to understand. Is how do you take the unstructured data that you're getting and that other people are getting and use that unstructured data to improve your business and also understand how other people are positioning their business.
[00:04:48] George Swetlitz: And so we have some features, we call them review sentiment analysis. Competitor sentiment analysis, competitor analysis, where we take all of this information and we help businesses understand. What they're doing well, where they can improve what their customers are doing well, and how they can use the areas where their competitors are faltering to better position themselves.
[00:05:13] Adrian Tennant: Traditional review platforms often provide a single rating that doesn't tell the whole story. How does your sentiment analysis approach differ?
[00:05:25] George Swetlitz: So the way to think about it is what we do is create mini ratings. So a review gives you a single rating. And reviews are very complex. I love the burgers, but I hated the fries.
[00:05:39] George Swetlitz: It was too loud. But other than that, the ambiance was great. There's a ton of information that can seem overwhelming if you decide to try to analyze it yourself, right? Coming through in all of these reviews. So we take the reviews, we break them down into phrases, and then we use AI to assign those phrases in their context.
[00:06:02] George Swetlitz: To topics that we develop for businesses, but that they can modify. So essentially what you then get is what we call percent positive mentions. So if there are a hundred mentions of food quality and 90% of those are positive mentions and 10% are negative mentions, you have a 90% positive on food quality, and that's very understandable.
[00:06:28] George Swetlitz: If you have a 4.8, well what's a 4.8? But if you have a 94% positive mentions, you know exactly what that is. And so what it allows you to do is say, I'm looking at my business and I see that my percent positive on quality is 70%, but my percent positive on service is 95%. So you know precisely what your customers are talking about and what you need to do to improve.
[00:06:55] Adrian Tennant: You've talked about how you analyze comments. Now let's talk about your platform's, AI powered Response Generation. George, how does Right Response AI create personalized responses rather than generic automated replies?
[00:07:12] George Swetlitz: I'll start by talking about a couple of observations. One observation is that you increasingly see people being frustrated with generic AI responses.
[00:07:25] George Swetlitz: Most review management platforms use generic ai, and they just essentially spit back what's in the review. And oftentimes you'll see people respond to that and say, you are just using a bot. There's nothing, you're not even looking at these things, and people get frustrated about that. The second thing is.
[00:07:50] George Swetlitz: And this is an observation that I had back at Alpaca. There's no one closer to the bottom of the funnel than someone actually reading your reviews. They are deciding and they're looking for a conversation. More people read reviews, then visit websites. So the challenge for a true marketer is to say, how do I bring my website to the review conversation?
[00:08:21] George Swetlitz: And what we've seen is that when you engage people, when you provide people with useful and helpful information in your part of that conversation. The customer initiated it through the review. You now have an opportunity to engage in that conversation, and you can respond by using a template which says nothing by using a generic AI, which says nothing, or you can actually respond to them in the way that you would if you were speaking to them face-to-face.
[00:08:53] George Swetlitz: That's what we do. So that's the context. How do we do it? Well, we have a whole series of AI agents that act on every single review that comes in. Some of them are simple. We look at the name, is the name, a name that we should use. In the response, the agent makes a judgment, and if it seems like a name, we use it.
[00:09:16] George Swetlitz: If it doesn't seem like a name, we don't use it. That's a simple AI agent. A more complicated AI agent is our fact system. So essentially what we do is we look through your last 500 reviews. We identify the things that people talk about in your reviews, and we create a fact. So let's just say that people complain about the fact that they can't find parking.
[00:09:46] George Swetlitz: And so we say if somebody says something negative about the fact that they can't find parking, tell them that the next time that the parking for the building is behind the building off of Beach Street. That's a helpful and useful piece of information that the AI can determine when it's relevant to a response and include it directly into the response, and so people can generate 10, 15, 20, 25 facts about the business, both on the positive side and the negative side, and those are seamlessly incorporated into responses.
[00:10:20] George Swetlitz: Allowing the business to participate in that conversation and provide a useful response both to the customer that left the review and to all the review readers that are trying to decide where to go next.
[00:10:35] Adrian Tennant: Okay. You've told us about the responses. How does your platform create personalized review requests?
[00:10:44] George Swetlitz: I'll start with the context again. The key with review requests is one, to get more reviews. You're trying to get as many reviews as you can, and you're trying to get more good reviews. That's the key, but the first part is probably the most important. Which is conversion of the request into a review. And what we found is that the more you can make an emotional connection with that customer, the more likely they are to actually leave a review.
[00:11:17] George Swetlitz: Everyone gets every day review requests, hi, we saw you yesterday. Could you leave us a review? What is much better is something that's more personalized. Personalized with things that you know about the business. Hey, this is Phil from The Roofing Company. We were really excited to complete the project yesterday.
[00:11:40] George Swetlitz: Here's a photo of the finished project. It really helps us get the word out when people talk about why they liked our service. I've included a couple of questions below that our customers value knowing about, because they ask us these questions all the time. If you wouldn't mind leaving a review to help us get the word out, we'd really appreciate it.
[00:12:02] George Swetlitz: You've personalized it with something real. Who was the person you were engaging with? What were they doing? When were they there? Here's a photo. Here are questions. You're much more likely as a customer to actually leave that review, and that's what we do. So we tie in to the CRM systems that have information about that customer.
[00:12:25] George Swetlitz: And we use AI to personalize that review request.
[00:12:30] Adrian Tennant: Well, as regular listeners know, we love case studies, Oncle focus. So George, can you walk us through a specific example of how the combination of sentiment analysis and personalization works in practice?
[00:12:46] George Swetlitz: We work with a large home services business.
[00:12:49] George Swetlitz: They're doing everything right. They generated lots of facts. They auto generate their responses, not auto publish, but auto generate. And then they distributed that responsibility out to the regional managers along with giving them another tool that we have, which is an AI generated monthly review analysis.
[00:13:10] George Swetlitz: So every month, the regional managers get a synopsis of what happened every day they can go in, they see review responses that have been automatically generated using the facts. Most of the time I would say about, based on our data, 85% of those responses, they just accept the rest. They'll make a slight modification and publish.
[00:13:34] George Swetlitz: When we first started working with them, they were at about a 90%. Percent positive, right? So good business, 90%, but they were only responding to about 20% of the reviews. Now, with no additional resources, they're at 95% positive. 80% response rate and they've told us their business is booming. Great example, we have a business that serves groups, so they serve individuals, but they also do groups, family reunions, other kind of large groups, those kinds of things.
[00:14:10] George Swetlitz: When they send out a review request, they send a photo of the group and they to remind them of that event that they did together, and then they ask for the review. And they've tripled their conversion rate, so it was sub 10% before, which is kind of a typical conversion rate. And it's gone up to about 30% for that type of, you know, request.
[00:14:32] George Swetlitz: Big improvement based on personalization of requests. Then we had another very, very large business, more than a hundred locations. Response quality was low, response rate was low, sub 50%. They had two people. They couldn't get to everything. They had negative reviews that were about 15% of the total, getting lots of reviews.
[00:14:53] George Swetlitz: They couldn't get back to everybody. They just couldn't get their jobs done. With us, they literally went to a hundred percent response rate. Not difficult to do with our system. 90% of the time that they spend now is spent on real issues with real customers, and they have one FTE. So that one person is now not focused on all of those reviews, but focused on the people that really need their help.
[00:15:20] George Swetlitz: So just ways that automation, sophistication, focus, and leveraging all of these tools can drive performance.
[00:15:30] Adrian Tennant: Yeah. Great examples. Let's take a short break. We'll be right back after this message.
[00:15:36] Advertisement: Hi, I'm Devin Bramhall, author of B2B Content Marketing Strategy, A Media First Framework that Accelerates Growth Published by K and Page.
[00:15:45] Advertisement: So if you've ever felt like your content marketing creates a lot of activity, but very little real business impact. This book is for you. I wrote it to help mid-career marketers move beyond outdated playbooks and build human-centered strategies that actually drive growth. Drawing on my experience as a former CEO of animals where I worked with brands like Google, Amazon, slack, and ramp.
[00:16:06] Advertisement: Plus real world case studies from companies like Cisco Help Scout and Nevada. I present you with modern marketing frameworks that combine community building with AI and multimedia based tactics to create growth strategy that gives companies unique advantage over the competition. As an unclear focus listener, you can save 25.
[00:16:24] Advertisement: Percent on B2B content marketing strategy. When you order directly from kogan page.com, just enter the exclusive promo, big I 25 at checkout. Shipping is always complimentary for customers in the US and the uk. I hope my book helps you create content strategies that build real connections and drive measurable results.
[00:16:45] Advertisement: Thank you.
[00:16:47] Adrian Tennant: Welcome back. I'm talking with George Switz, co-founder of Right Response ai. George, how does Right Response AI partner with marketing agencies? I'm curious what white label offerings might look like.
[00:17:03] George Swetlitz: So we do have a white label program and essentially the agency has all of the tools at their disposal.
[00:17:12] George Swetlitz: We have a very different pricing model. You know most people in this space, you have to decide, I'm gonna buy the request package, or I'm gonna buy the review package or the map rank tracking package. We don't do any of that. We have a per location fee, which if you buy annually is $8 a month, $10 a month if you buy monthly.
[00:17:33] George Swetlitz: So it's very cheap per location. And then everything above that is usage based. So if someone is focused on review requests, they can come in, start with a review request and just pay for review requests. If they have a location that doesn't have very many requests, they don't pay much because they're not using the system that much.
[00:17:52] George Swetlitz: And then over time, as they build their business, they're paying more because they're using more. So it's a very different model and it respects small and large companies as well. There are discount packages as you purchase more credits and all of those kinds of things. And so agencies get another benefit, which is agencies can use our software for research and they don't pay a per location fee, so they can go in, they can scrape any site they want, any location they want, they can do sentiment analysis and just pay.
[00:18:27] George Swetlitz: For the sentiment analysis that they do. So if they're getting ready to do a pitch, they can go and they can download the reviews for 10 locations and 10 competitors and analyze all of that as they're prepping for their pitch. And it costs them, you know, what, 20 bucks to do something that they might have an, you know, a junior person spending a week on.
[00:18:49] George Swetlitz: And so the agency just simply takes over. It's under their brand and. They can use all the features we have.
[00:18:58] Adrian Tennant: Perfect. Now, back on the client side. For executive teams working with franchise or multi-location operations, how can they utilize review data to identify operational issues before they escalate into much larger problems?
[00:19:17] George Swetlitz: Yeah, we have one chart in particular that we developed that's exactly for that purpose. And so essentially what it does is you can decide which locations you wanna put on the chart. You can put locations on the chart, you can brands on the chart, you can put regions on the chart, whatever you want to do.
[00:19:36] George Swetlitz: And it compares visually their percent positive by topic. And so I'll often sit down when we're onboarding large clients and show them this with their own data and say, look at this location, A, they're literally underperforming on every single topic. You think you have an issue there and it just opens their eyes because all they had before was essentially, well, the average rating for this location is 4.4, and our average is 4.7.
[00:20:12] George Swetlitz: It's much more tangible. When you say, look, their percent positive is in the seventies in every single topic. I think you need to talk to the manager. So that is how, that is the primary way that large organizations use our tool to get at those differentials.
[00:20:31] Adrian Tennant: Right now, your platform is designed for businesses with physical locations.
[00:20:37] Adrian Tennant: Do you have any plans to extend it, say as a plugin for Shopify online stores? Asking for a friend?
[00:20:46] George Swetlitz: Yeah. That that's coming. You know, there's just so much to do. You know, technology is changing so fast and it opens up so many opportunities that we always have to decide. Do we go wider or do we go deeper?
[00:21:01] George Swetlitz: And right now we're in the deeper phase, but eventually we'll go wider.
[00:21:06] Adrian Tennant: Mm-hmm.
[00:21:06] George Swetlitz: But that's not something we have right now. Now, I will say that many companies that are not necessarily location based. Are on Trustpilot, for example, or Better Business Bureau or those kinds of platforms, and we support those 100%.
[00:21:23] Adrian Tennant: Got it. Well, it's in the name you've built, right? Response AI entirely around artificial intelligence. So George, what have you learned about implementing AI effectively in business applications?
[00:21:41] George Swetlitz: So I would say there's two key learnings that we come back to again and again. One is that when you're working with ai, smaller is better.
[00:21:52] George Swetlitz: The more you try to put in a single prompt, the more likely it is to fail or be erratic. We just had this yesterday. We continue to develop our approach to generating facts for people, right? So when they come in, we look at 500. Reviews and we developed the facts, and that's a lot of data, 500 reviews and you're trying to cluster them.
[00:22:16] George Swetlitz: There's a lot going on and it was failing even with the best models. It was failing. We kept on breaking it into smaller pieces. And then we realized that we were asking it to do math and language models don't do math. And so it was failing on that problem, right? So we had to take the math bits out of the prompt and do that on the backend.
[00:22:42] George Swetlitz: My point is, the less you can do in any single prompt, and the more you can make sure that the prompt leverages the strength of the ai. The better off you're gonna be. That's, I would say, one thing and the second, and I sort of alluded to it, is doing thinning in pieces. Kind of chain of agents as opposed to trying to develop one prompt that does a lot.
[00:23:09] George Swetlitz: We have many, many, many agents that hand off one to the other, and we get much more consistent results when we do it that way.
[00:23:18] Adrian Tennant: Relatedly, thinking about hospitality and multi-location businesses, what's your perspective on striking the right balance between automation and maintaining authentic customer relationships?
[00:23:35] George Swetlitz: That's a great question. Often when I will have an initial conversation with a company, one of the things that they'll say is, we're worried about, you know, this. Fake personalization, and what I'll say is I agree with you, but that's not what we do. We don't do any fake personalization. I'll give you an example of fake personalization.
[00:23:58] George Swetlitz: I got an email, so you know, I'm a B2B company. I get a lot of B2B unsolicited emails and that's fine because sometimes I get one that's interesting, but I get a lot, and I just got one the other day that said, oh, I see you went to Penn undergrad. You know Penn's a great restaurant town. Have you ever been to this restaurant?
[00:24:21] George Swetlitz: Well, just for fun, I looked it up. That restaurant only opened maybe 25 years after I graduated. Okay,
[00:24:28] Adrian Tennant: great.
[00:24:28] George Swetlitz: So to me, that's fake personalization. It's just AI looking for something about me to strike up a conversation completely fake. We don't do that in our responder. It's not fake personalization. The business is deciding how do they incorporate their values, their company.
[00:24:48] George Swetlitz: Into a response that's real personalization. It is bringing their business to the customer in our review request. They know it's their customer. They know things about their customer. They're bringing that to the request. That's not fake personalization. That's real personalization. So I would say I'm very afraid of faith personalization because it can go wrong.
[00:25:12] George Swetlitz: Like the example I just gave you, but real personalization is based on real things. It's controlled by the business, and it just relates to the fact that the people who know things can't be everywhere at every minute. And AI gives them the ability to extend themselves in a very useful way. Hmm. Agreed.
[00:25:37] George Swetlitz: Well, looking ahead, how do you see the role of customer feedback evolving in marketing strategy? People want to be heard. More and more people want to be heard. They want to be engaged with. They think that should happen quickly, and they write. Reviews are going up and up. The quantity of reviews can continues to go up, and businesses have a question of it to answer.
[00:26:08] George Swetlitz: Are we going to try to do good things with all that information or not? And so I think what will happen over time is that companies will realize. That if they engage with this review ecosystem in a sophisticated way, they will win over companies that don't. We see that with our clients. When they engage in a sophisticated way, they win.
[00:26:33] George Swetlitz: And so I think as more companies see that and more companies realize this isn't just a task. This is a conversion engine, they will engage more with it.
[00:26:47] Adrian Tennant: Hmm. Great conversation. If listeners would like to learn more about Right response to AI or connect with you directly, what's the best way to do so?
[00:26:59] George Swetlitz: So the best way is through our website write response ai.com.
[00:27:05] George Swetlitz: We have this ability there to schedule a a video conference. And if you put in the notes, I wanna talk to George. You'll get me if people wanna follow what we're doing, write response. AI has a LinkedIn page and people can connect with me there as well.
[00:27:22] Adrian Tennant: Perfect. George, thank you very much for being our guest this week on In Clear Focus.
[00:27:29] George Swetlitz: Great conversation. It was a pleasure talking with you. Thank you.
[00:27:32] Adrian Tennant: Thanks again to my guest this week. George Switz, co-founder of W Right Response ai. As always, you'll find a complete transcript of our conversation with timestamps and links to the resources we discussed on the Incl focusPage@bigeyeagency.com.
[00:27:50] Adrian Tennant: Just select insights from the menu. Thank you for listening to In Clear Focus, produced by Big Eye. I've been your host, a Adrian 10 Until next week. Goodbye.