How Respond.io increased trial clicks by 162%

Arthur (Head of Community & Customer Marketing): Alright, here we go — another Finer Demos Club session. We'll give it a little bit of time as people come in. Feel free to let us know where you're calling in from. I see Mark — you're eager for this one, I'm glad, because this might be one of the best ones. Awesome, Kansas City. And, Nabila, did you want to share where you're calling in from?

Nabila (Product Team, respond.io): Sure. I'm actually calling from Malaysia, which is in the Southeast Asia part of the world. Whereabouts are you, Arthur?

Arthur: Oh, I never told you this. Right now I was in Las Vegas yesterday for the Dolby Summit. I'm here for about a week, because the Forrester Summit is next week, Sunday to Wednesday. Shout out to the UK — Belfast, Ireland, love to see it. We have quite an international crew here today. Atlanta, we always gotta represent.

So this one I'm super excited about, because as we were chatting through, Nabila and her team had such an interesting way of iterating. We also want to test out something new for this format: at the end, we'll be doing a quick little workshop on the A/B testing feature. So you can learn from Nabila's team everything they've done, and then see how we try to make it a lot easier for you, so you don't have to go through as many iterations. No way — Doug Fisher, that's amazing. Fun fact: my first tech job ever, Doug Fisher hired me, from Lone Wolf. So that's amazing to see. I'm guessing you're a customer now — we have to catch up after this.

Derek, here's what we always say: we are recording it, but we don't make it publicly available. If you have to drop off at some point, the fact that you all registered means you'll get a private link to see the recording. I know this was a bit of a different time zone, because we have Nabila from Malaysia, and I know it's very late for her. So let me start sharing my screen. Before we get into the good stuff, Nabila, if you could introduce yourself, let us know what you do at respond.io, what respond does, and tease a little about what we're going to talk about today.

Nabila: Yeah, sure. Thank you, Arthur. Hi, everyone. So lovely to meet everyone, and thank you so much for taking some time to hear me share my experience with Storylane. My name is Nabila, and I'm part of the product team at respond.io. respond.io is an AI-powered customer conversation platform. We help businesses around the world manage all their customer messages — whether from WhatsApp, Instagram, Facebook, TikTok, website chat, or email — all in one place. We have AI agents that handle conversations automatically, so your human team only steps in when it really matters. Today I'm going to walk you through how we went from a demo with Storylane that most people didn't finish, to one where our trial clicks went up 162%. I'll share my tricks — the stuff we did so you don't have to repeat the same mistakes, what didn't work, what worked, and honestly what surprised us the most. So I hope that teased you guys a little bit.

Arthur: Amazing. So, I know our audience is going to be looking at this slide nodding their heads — this is the classic product-demo dilemma, so much to show. Walk us through what you were up against when you first started with Storylane.

Nabila: Yeah. We serve a wide range of industries — ecommerce, health care, travel, you name it. So every visitor that comes to our website or home page is there for a slightly different reason. And before we even had an interactive demo, there was really no way for anyone to actually see or experience the product before signing up for a free trial. And even if they did sign up, they only had seven days, so that window is relatively short for them to really get to the "aha" moment. So the risk we had, when we were trying to figure out how to build this demo, was finding that balance — sharing everything to show how amazing we are, but then maybe ending up not communicating anything. That's exactly why we started to dabble in this whole interactive demo with Storylane.

Arthur: We're going to go through all of these stages in depth, so don't worry, we'll come back to this slide. But this gives you an idea of how much thought went into all of their iterations, the breakthroughs they had at each stage, and what changed along the way. So let's jump into v1. Kick us off with what the first version started to look like.

Nabila: Oh my god, v1. I'm seeing the GIF now, and it feels like a while ago, but it's only been about a year-plus. Our version one took about a month, from blank canvas to going live. And even getting that first version out wasn't a one-person job — we had at least four teams. We had the product team, who defined what our product USPs were. The marketing team looked into the copy and what language works well for the home page. The sales team verified the language in terms of whether it matched what they actually hear on sales calls, which was super valuable. And last but not least, management was part of the review, because they speak to a lot of people and have insights and contacts we didn't have. And, of course, the Storylane CSM we had — oh my goodness, he was incredibly patient with us. I was very annoying when I started out because I was so impatient. We had so many questions, and it was so nice of him to attend to us.

I think the format we went with for version one was the biggest decision we had to make first, because we were like: should we go short, with one chapter, or go long, with three? We were nailing two or three core USPs at that point. So, long story short, we went with three, because it felt really thorough and comprehensive — let's show three chapters with our main core modules and see how that goes. So that's our first ever version.

Arthur: It's definitely the most "paralysis by analysis" — what do we show, all these questions. But you all came down to this incredible iteration engine. It's going to sound intense, but even as we were talking, you're like, "No, it was actually pretty scrappy." So walk us through how this actually worked in practice. What did reviewing the numbers every two weeks really look like in the day-to-day?

Nabila: I know that sounds like, "Oh my god, every two weeks reviewing numbers," and "iteration engine" does sound a bit formal and serious. But honestly, it was mostly about accountability, because every one of us knew this demo was sitting on our home page and everyone was watching. There were three things. First, we really looked into what we track — we nailed down what metrics we're actually seeing the progress of. So we have engagement rates, demo completion rates, the average timestamp, drop-off points, and our CTA click rates — demo clicks and free trial clicks. We monitored them every two weeks. Why two weeks? We thought one week was too short and three weeks was too long, so, two weeks. Pretty scrappy, but it seemed to work for us.

Once we knew what to track, we then asked ourselves: where exactly are people leaving, and what's the drop-off? Why are we losing them? Is it the content? Is it the format? Are the right people clicking the right CTAs? Does the current version actually reflect the most important value in our product? So we were constantly questioning and evaluating that specific version, and mapping out whether it actually suits our business goals and our product-market fit. And once we knew what we track and what to ask ourselves, we looked into recommendations — what do we do after? We cut pages that were repetitive or where people dropped off, or that were too process-heavy. We looked at restructuring the chapters — maybe people are dropping off before the main highlights, so how do we change the story? And then running A/B tests, because, like I said, it took a village to get through a version. There were a lot of opinions, a lot of people saying, "This doesn't make sense, I don't like this, I like that." So I think A/B testing is the way to go — experiment, what works, what doesn't, because we were all doing this for the first time. And we played a lot with design and localization. So it was a lot of trial and error, every two weeks.

Arthur: It sounds like a lot, but you've broken it down. One of the cool things that came about almost by accident while we were chatting — you said, with your specific use case, this was a unique bonus, because in terms of working cross-functionally, you didn't necessarily have a formal project.

Nabila: Honestly, it was a no-formal project. I remember the day I got assigned to this — they were like, "Oh, we should do an interactive demo. Oh, you know what, Nabila should do it." And I was like, "Wait, why? What's going on?" There was no formal project, no big recurring meetings. I just pulled in who I thought I needed to involve. And what made it work, personally, was the reporting and the visibility — because every two weeks, everyone got the same report, same data, same context, same recommendations, and we aligned on all the metrics. Every time we put in these reports, we included a deadline, so people felt informed, not forced, not surprised. That built trust across all the teams. Over time, our stakeholders became more comfortable with "we don't know yet, let's try," because that visibility gave a bit of certainty that our team knew what we were doing, and that we were trying our best to always improve the demo, all data-driven.

And something cool happened — the demo became a shared project that brought all our teams together. It was an unexpected side effect, but very positive, with that culture of trial and error, because, you know, we never know if we never try.

Arthur: One of my favorite quotes from the best Pixar movie ever, Inside Out — they comment that people always get opinions and facts confused. So I love that people had a lot of opinions, but as you went through, it's like, "Hey, if we have an opinion, let's test it against the data." I know we'll have a Q&A at the end, with the workshop, but I thought Alex's question is on par with what you just went through. He asked: what volume of demos do you need to reach statistical significance? Was it more about the two-week timeline, or were you looking for a number for statistical significance?

Nabila: We did look for statistical significance. For the A/B test phase specifically, it wasn't a two-week monitoring process — we did monitor every two weeks, but we didn't report the progress of it, because with the statistical part, you have to get enough sample. So we let the experiment run for about one and a half months. And if I'm not mistaken, we got at least about 2,000 per demo, and we had three demos that we were A/B testing.

Arthur: That's awesome. Great question, Alex — I'm sure you have some follow-ups we can answer at the end. So now let's get into the depth of all these different versions, what they look like, what you learned.

Nabila: Okay. This is a really small recap of everything that happened, so I won't go into too much detail — I'll just cover the main highlights. Generally, every version had different types of changes: small stuff like copy changes, page changes, content changes, design changes, but also some very big, major changes. I'll walk you through the more relevant ones.

So, version one. It was truly an experiment as a whole, because we didn't know what to expect — we were seeing numbers we'd never seen before, so we were truly excited when we launched. We're like, "Wow, people are engaging." But when we looked back, it was 30% completion, with 70% actually dropping off before they get to the good stuff. We thought people would go and click through the chapters, because everything was so clean and structured and people might want to come back and learn. But we were quite wrong — a lot of people just engaged with the first chapter. So we were like, is it the duration? Do the chapters not appear? Storylane did have a lot of features we could customize, so we played around with those in our mini-iterations. This was a good start, and we used it as our baseline.

Arthur: Doug, I think Nabila might get into your point here. I know this was a big unlock for you, going from v1 to — let's jump to v3. Talk us through some of the changes you made here.

Nabila: For v3, it was a big jump, because it was a feature I was playing around with in Storylane. Every two weeks you're in the platform, trying things out, looking at the metrics, and I saw the new release they did — the video walkthroughs and the voiceover feature. I was like, "Oh, this would be interesting, maybe we can try this out." We floated the idea with our management — "Maybe we should humanize it a bit, maybe that could help with completion, because we don't know what we don't know" — and they were like, "Let's go." So we had our lovely head of sales walk us through it. She's so funny — I really like the way she talks — so I was like, "Hey, you're doing the voiceover." For this version, it was a bit more planned. The guides, the text, were more action-oriented, and Jess, our head of sales, in the video talked more about the outcome and the value and the benefit of that specific action. We thought, "Wow, this is amazing, the user gets the full experience." Our engagement rose to around 61%, and CTAs also went up 20%. But it introduced a whole new scaling problem, because Jess is only speaking in English. How do we convert Jess speaking into all the other languages we support? We have a lot of customers from Spain, from LATAM, from EMEA. So that brought up this whole conversation about a problem we'd observed.

Arthur: So we're at v3, and now we're jumping to some of the final versions. Now you're live across multiple channels. Talk us through what happened across the latest ones.

Nabila: So the localization problem started to make us really question ourselves. Were we making the right assumption with the scaling problem? Is the format right? So we did an A/B test, just before v6. Version A was Jess speaking — voice, video, and text. Version B was just voice. And version C was just text. Because there were so many opinions against it, I was trying to find a way to get people to be more conscious of our decisions based on data. And we realized we were completely wrong, because what our audience wanted was actually text. So text won, everyone. I know, right? You would think we'd have won, but for our audience, they actually prefer text. We realized a lot of them liked that self-paced demo — they could scroll, click here and there. That resonated most with our audience. So we were totally different from what we thought customers wanted. So — A/B test, guys.

So we totally scrapped the three-chapter structure entirely. This was the first time we did the big jump to our latest version: we only have one chapter, one focused story around our core value proposition, which is the omnichannel inbox. And we had AI agents at that point — we'd just launched it — so we made that our main feature. Completion rate shot up from 42% to 64%, and with that, trial clicks went up 162%, because people were able to complete the demo a lot faster, see the "aha" moment faster, and we were able to communicate the value of the product faster. So, thanks, Arthur — sorry, I jumped the gun there.

Arthur: Oh, no, don't worry. This was fantastic. It was such a cool way to see how much you iterated, how much you A/B tested, and came into new problems like localization, especially being a company that sells so globally. Walk us through some of your biggest takeaways as you were getting this up and running.

Nabila: There were a lot of takeaways. The biggest one is definitely starting with the story, not the features. As a product person myself, we get very fixated — "Oh, this feature is so cool, we should feature this." But actually, nobody wants a product tour; they want to see how their problem gets solved. The feature is just the "how" — they want to know the "why." So before you jump into creating a demo, focus on the story you're trying to convince your audience of. We spent a lot of time on the storyboard. That's number one.

Number two, in terms of metrics: completion rate is your most honest metric, for me, and trial starts. Impressions lie. Engagements can be accidental — they can be traffic-driven. But completion, and the trial clicks, tell you if your story has helped and if they might actually convert to customers. So choose your metrics wisely. Of course, it depends on the purpose of your demo, but for our case, completion and trials were the primary metrics we were most focused on.

Third is formatting — never assume a format. I thought Jess speaking was why our completion rates went up; I loved the way she presented. We were convinced, but we were wrong. So don't assume it, just test it. Let your audience tell you, because every business and industry is different.

And last but not least, build for iteration, not perfection. The perspective we had when we started: we knew version one was never going to be the best version. Internally, all of us said that — although they never said it to my face, I know everyone was like, "You need to improve this." But I'm more proud of the fact that we have a system to review, analyze, and improve as we go. You always have to have that perspective — to keep improving as the product grows and your audience changes. So those are the four key takeaways for us.

Arthur: That's fantastic, Nabila. As we drum up some questions — let's get to Doug's, because he's been patiently waiting. He mentioned the CTA specifically — book a demo versus free trial. He asked: does this demo have a CTA to allow users to self-serve the trial, or is it "talk to sales"?

Nabila: Yes, we have two CTAs. One is "book a demo," which leads to talking to a sales team. But the "start free trial" one directs them straight to being able to self-serve, create an account, do their trial.

Arthur: And is that the main one you focused on within Storylane, or did you test both?

Nabila: We tested both, but we eventually shifted to focus more on the self-serve — the "start your trial" one.

Arthur: Amazing. As you can see, so many iterations — it's fresh in Nabila's mind. So think about some questions. And as you're thinking, we actually want to bring up Renee, who's going to show us how to use the A/B testing feature. As you go through this, start taking Nabila and her team's frame of mind and thinking about your own demo, in this framework of: "I think changing X will improve this metric, because of this reason." Just to get you thinking — "I think adding a CTA earlier will increase trial clicks," or "I think a shorter demo will improve the completion rate," or "I think removing chapter two will reduce the drop-off." So, Renee, would love for you to take the reins here.

Renee (Storylane CSM): Hey, guys, hope everyone is doing great today. Excited to have everyone on — such an awesome use case. The A/B testing in-platform was definitely something we had a lot of requests for. I'm going to walk you through where to find it, how to set it up, and then we'll check out a completed test from our own home page, so you can see the A-to-Z of how it works.

So you'll pop into your Storylane dashboard, and then under Analytics, you'll find the demo you want to use for the A/B test. Let's assume it's already up and running. When you click in, before it was just the aggregate view of the analytics, and then you'll see some tabs up top for the A/B test. Once you hit "create the A/B test" — this is going to be a gated versus ungated demo, so you can imagine what metrics we'll see. You'd have your variants: this is A, which is the one we're going to duplicate and slightly change — in this use case, gated versus ungated. We recommend having the naming convention be the overview, but with something in it that's "variant B" or "version B." You pick the demo, and you put the traffic split — fifty-fifty is what we went for here. Once you click "create test" — we're going to pretend it was on for a bit — this is what you'll see: how long it's running for, which one actually won, and then the comparison down here for CTA clicks, completion, leads, intent, time spent, and the status. So B was the winner here, which you can see. And below, on the graph, it gives you a day-by-day of how each demo was performing. It makes it really easy to do this all within the platform. So if you have any questions, please reach out to your CSM or us in the chatbot.

Arthur: Rachel had a question — she's saying, is it available in the starter plan, the A/B testing feature?

Renee: I believe it is. I can definitely double-check. I'm more than happy to trial it for you — if you'd like, I'll get your email and follow up. If it's not, I'm happy to give you a quick trial of it.

Arthur: Amazing. Doug, I'm loving the questions — we definitely have to catch up after this. He asked: other than the demo on the website, what other channels do you distribute the demo on — LinkedIn, Google Ads, other? And what's been most successful? So, Nabila, talk us through the channels you've distributed your Storylane demos on.

Nabila: Our current demo is primarily for sales, and more for the top-slash-mid of the funnel. We focus more on the home page, we have it for our G2 and review sites, we have it for referrals, and also in conversations when salespeople are speaking to their customers or leads. And which one has been most successful is definitely the home page. I say that because right now we have a huge surge of self-serve customers, and the home page demo is the one driving them mostly, without external reviews or referrals or a sales assistant.

Arthur: You just answered my follow-up — there's a sales-assist motion where you're looking at free trials, how they're getting through, and then somebody would reach out to book a demo?

Nabila: Yeah, yeah.

Arthur: Perfect. One thing you said — referrals. I'm curious how you use it for referrals.

Nabila: So we have a referral program. Any of our loyal customers can refer, and if they have a friend or colleague who then subscribes to respond.io, they get some form of rewards. So they can use our Storylane demo as a resource to pitch — "This is how the platform looks, this is what you can do" — assisting that referral process for our customers to get that reward.

Arthur: Oh my gosh, that's genius. I've never heard of that, so I'm glad I asked — and a great question, Doug. Edgar, good to see you again — I believe you were quite active at the last Finer Demos Club session. This might be a question for Renee: is there any development on A/B test hubs, or exporting A/B test data?

Renee: Hey, Edgar, I was just typing back to you. Not at this time, but we'd love to — we'll add it to the roadmap. I'd love to hear in chat, or we can follow up, on how you'd want to use the hubs in the test. Is it for sales outreach, or something that lives on the site? Any additional information would be great for our team to have. I'll add those to the roadmap, or ask the product team.

Arthur: Edgar, keep them coming — this is one of my favorite things in the CS channel. Customers will ask for features, and the team will say, "Hey, we don't have that yet," and then sometimes later that day they're like, "Actually, looks like we just released it." So keep them coming, because our team is always trying to get feedback on where we can iterate the product. We've got Travers — hopefully I'm saying that right. At a high level, what is A/B testing? Is it one link, but users get randomly directed to one or the other? Renee, do you want to take a stab at that?

Renee: Yeah, for sure. With A/B testing, essentially it's a version of a demo that you want to make minor tweaks to, to see which has the best engagement, the best leads. The example I was showing was gating versus ungating, which is a really great use case. And you have the ability to put the percentages in — you can do fifty-fifty, eighty-twenty. We'd probably recommend more of a fifty-fifty. So 50% of users would be sent to one demo, and 50% to the other.

Arthur: And, Nabila, from your perspective — I don't know if you got a chance to use the A/B testing feature later once it was developed, but maybe speak to how you were thinking about A/B testing, whether you were sending 50/50 traffic to each iteration.

Nabila: Ours was three versions, so it's like A/B/C testing — thirty, thirty, thirty. And I think one was 40. The baseline, the version we already had, was the 40, and the two versions we hadn't tested were the thirty and thirty. But ideally, fifty-fifty is possible.

Arthur: That's great. Rachel, I think we're on the same wavelength — she wanted to hear, do you have a separate demo you created for referrals specifically, and how does that differ from your other demos?

Nabila: Yes, we do. For our home page demo, the first step has a slightly different intro, and the referrals one has a different intro, because we're just trying to speak to the right person. But in terms of the actual steps and the format, they're exactly the same. They're different versions, though, meaning the link we embed on the home page versus the link we embed in our referral materials are different, because we want to track the performance of the referrals and separate that from the home page.

Arthur: Awesome. Looks like Sandra has a feature shout-out: "A feature I'd love to see is the ability to quiz people on how to use our product by making the hotspots invisible — so you'd say 'start by clicking on the play button,' and they have to find and click it themselves without the hotspots showing where it is."

Nabila: Yeah, they'd love it. That'd be quite cool — I'll try that.

Arthur: Keep them coming, Sandra, this is great. We got one from Rebecca B: have you used the AI demo creation when creating demos in Storylane? Rebecca, give us a little more about what you specifically meant by the AI demo creation feature, but, Nabila, feel free to speak to it.

Nabila: Generally, on Storylane, we did dabble in the AI video avatars, the voiceovers, and also the translations — primarily because of that localization we wanted to address. The video avatar we played around with because, like I said, with Jess's videos, it was sustainable but not scalable, in terms of getting her to speak in all the different languages. We did try it, but at that point, we all loved Jess so much — my CEO kind of vetoed it, he was like, "I don't want anyone but Jess," and we were like, "Okay, figure out something else." I actually really enjoyed working with it; it sounded exactly like what we wanted. But it was an internal decision.

Arthur: Renee, anything to add there on the AI demo creation?

Renee: Yeah, two parts. For Sandra's feature request — you actually can make the hotspots invisible. You'd just make the opacity zero, so it's fully invisible, and then take out the text inside the guide. In the preview, it'll still show the little box, but when you go to actually share the demo, it'll be gone. So you could have an intro, like a media modal that says "start by this," and then have it autoplay the next step, which they'd have to figure out where it is. You can also add a time trigger on that one, so if they don't find it, you can help push them along to the next step. But the opacity being 0% is a nice little workaround. It'll be more visible in v2 as well.

And for the AI demo creation: when you're creating the demo, you have the option, in the screenshot or HTML, to toggle on "create with AI." This adds content with AI — you can change the language and the tone. There's "marketer" and "sales rep" for the different tones. I love using these. What I like to do on top of this — once you're in the editor tweaking the guides, you can use "Improve with AI" and put in another prompt. My prompt that works really well is, "Keep it engaging and value-driven," which goes back to: don't make it just "step one, step two." You can say, "This is how the product will benefit your company and your customers," which resonates a lot more, especially for top-of-funnel awareness. The more specific you get, it does a really good job at revamping the entire demo. So I'd recommend the creation, and then tweaking it a little more when you're inside the editor with "Improve with AI."

Arthur: Interesting. I think this is maybe what Richard is asking: is there a way to give Storylane context for the product? The AI text captioner is super helpful, but it makes some assumptions we could correct with a more detailed prompt or background context.

Renee: Yes, that's the "Improve with AI" inside of the editor — it's in the top right, a little pink button, and that's where you can put the prompt in. I make the Office-themed demos for our knowledge base, so making Storylanes with Storylane, of Storylane, for Storylane customers, gets very meta sometimes. I'll say, "Use emojis and the Office language," and it does a pretty awesome job. So if you see any of the Office Storylanes in the knowledge base, I probably made them, using "Improve with AI."

Arthur: Did you want to pop in and show Richard where to find that, in case he wasn't aware? And as Renee's doing that — I know we have about five more minutes, so any other questions for Renee or Nabila about A/B testing, feel free to type them in the chat.

Nabila: I really like your Office-themed campaign, by the way. I love it.

Renee: I'm so glad. No one talks about how amazing the Office is — sometimes it's too underrated.

Arthur: We love it here, I will say. Haven't gotten in trouble yet, so hopefully we won't.

Renee: This is a chapter-menu demo I made. You can see it already has an Office GIF here. This used the AI-enhanced demo. This is going to be the new UI, but here's where you put that prompt in. There's also some inspiration, with different funnels — marketing, lead gen, top of funnel — but you're not limited to these, and you can put a lot more on top. Let me do a fun one: "tailor it to Michael Scott, and keep the demo engaging and value-driven and fun." Again, these are for the knowledge base, so I have a bit more wiggle room. You can also apply it to multiple chapters — pick and choose, though you'd probably want to do all of them. And you can have the AI voiceover added along with it; it'll be translated with the new guides. So if we click "enhance content," it'll wait a few seconds — the demo's a little longer. It'll have you go through and accept the changes, or cancel them and try again. Best practice here is to duplicate the demo before you start playing around with the AI, because once you've accepted it, you can't go back. So I always duplicate it and play around there.

Arthur: While we're talking about the Office — I don't know if we've announced this officially yet, but season three of the Demo Dundies, which is our Office parody, is coming up. The one Ranga shared in the chat, of Avix — they were one of the winners of season two. This is a fun celebration we have of our customers, with different categories you can submit. In the past two years, we've sent folks to Paris and Rome. So we want to continue the Office theme — submit your demos, because we want to see them and potentially use them as examples, and it could win you a fun little trip. Go ahead.

Renee: Yeah, I still have my "Serenity by Jan" little candle here. I printed out, like, a hundred of these stickers back in January last year. I think Arthur may steal my thunder this year and host it, but I'm really excited for the Demo Dundies. One of our customers went to Paris and had the best time, got engaged by her roommate, so it was awesome to see it all unfold on LinkedIn.

So, this is what the AI came up with — we can click "let's go" and go through the demo. I also forgot to mention you can edit specific steps too. So if you go through and say, "I like this one, but not the other one," you can edit or revert to the original. Right off the bat, that's a great start. There's going to be a lot of "that's what she said" — it's great, we love that. It did a really great job tailoring it. This is for Michael Scott, but you could say "tailor to a CMO, a CFO" — you can do a lot of different personas inside of that as well.

Arthur: That's fantastic. Well, it looks like the questions have stopped. Nabila, thank you so much for this — I really enjoyed hearing how it brought you and your team together, and how you tested and iterated. You've given the audience a really great framework to think about. And Renee, thank you for showing the A/B testing feature, so people can follow in the footsteps of Nabila and the respond.io team and start testing out these features. We're going to be doing another Finer Demos Club in May, so keep an eye out. Otherwise, thank you both so much, and thank you everybody for tuning in.

Nabila: Thank you.

Renee: Thanks, guys.