Conversion Rate Optimization with Microsoft Clarity’s Ravi Yada

Andrew spoke with Clarity’s Product Manager, Ravi Yada to understand exactly how Clarity can benefit your business and help you increase conversion rates.

Learn how it compares to Google Analytics, how Clarity can help identify website issues, and how to use it to check the rendering of your website across all platforms

Also available on:
Apple PodcastsSpotifyGoogle PodcastsAmazon Music
or from your smart speaker.

Andrew Veitch: Welcome to The Joy of Marketing with me, Andrew Veitch.

Super excited to be learning more about Microsoft Clarity today. Joining us from Seattle, we have Ravi Yada, who’s a Clarity product manager at Microsoft.

Welcome to the show, Ravi.

Ravi Yada: Thank you so much for having me, Andrew, I’m super excited to be here.

AV: So can we start by maybe just by telling us where the idea for creating Clarity came from?

RY: Sure. So Microsoft Clarity is a user behaviour analytics product to help you kind of understand the behaviour of users on your website. And as you can imagine, the reason it was created is basically because we had a need for it internally.

If you think about our product, Microsoft Bing, it has millions of users worldwide, using it to search for experiences.

And a lot of teams do A/B testing, building new features, improving existing features, and when we look at metrics and scorecards, it’s hard to understand what happened. Why did these metrics move in a particular way? So everyone wanted to understand the why behind what happened. And from there, the need for a behavioural analytics tool kind of developed, which ended up being a public product that we know today.

AV: Yeah, and I guess, if you’re dealing with a site like Bing with huge amounts of traffic, you can’t just sign up for a seven day free trial on a screen recorder where you just completely overwhelmed it, so I can see why you would need something pretty powerful.

RY: Exactly.

AV: So how does clarity compare with Google Analytics?

RY: Great question. That’s probably one of the first questions we get when we talk to some customers. Google Analytics is a great and very popular tool that helps website owners understand traffic analytics, right? Who’s coming to your site, where they’re coming from, where they visit, where they visited on your site, how many clicks do they do and some other useful metrics like bounce rate.

But what we say is, that’s kind of half the picture, you know, it’s like going to Yelp and seeing that a restaurant is, you know, good or bad without actual reviews of why people felt that way. And that’s where clarity comes in. Right? It’s a complementary tool to Google Analytics. And it improves your understanding of what users did on the page.

For example, you might see 20 clicks happened on your page, and you might celebrate. But clarity can show you what those 20 clicks were about and why they happened, and it could be that the user was frustrated, and trying to close this pop up, and ended up being frustrated enough that he had to or she had to to click 20 times. So clarity is really about showing the why behind the metrics, right, seeing the numbers as one thing, and then seeing the journey behind the numbers is another. So that’s why we believe it’s a very good complementary tool that you can actually use together, and many of our customers do use them together.

AV: Yeah, you know, I can kind of see that because when I was back in e-commerce, what I would sometimes do is actually sit down with a user and get the user to, to go onto my website to try and buy and I would sit beside them to see what they were doing. And I guess in a way, you’re doing the same thing, but doing it at scale.

RY: Yeah, exactly. And it’s really hard to find all those corner cases on your site to improve it, and that’s why having real traffic and real user insight really benefits that beyond just like, you know, a case study or user study.

AV: Yeah, I guess just on a security point of view, if you are recording what people are doing and typing into the website, isn’t that gonna be a problem with people, you know, typing a credit card number or something like that?

RY: Yeah, great question. So generally, Clarity, it doesn’t want to have any PII data centre clarity, and we aim to not capture any of that as well. So by default, we do mask any form input fields.

For example, we don’t know whether this is a search input field or user name or password or credit card, as you said. So we just mask all of those by default. And when we say masking, we actually obfuscate that content on the client itself, so the data is never transferred to the server. And then on top of that, we let owners decide their own masking amount. If they want to basically mask out all imagery in text, they could do that, or kind of be a little bit a lot more liberal and see more of the text on their site, depending on what their site contains.

And on top of that, you can customise with CSS selectors, right? Maybe there’s particular parts of the page that contain sensitive information and you can just obfuscate that and mask it out with clarity, CSS selectors, and then you won’t be sending any of that data to clarity.

AV: Great. Again, thinking myself as a user, and obviously I you know, like I’m sure we all do I buy a lot through e-commerce. You know sometimes I maybe go to an e-commerce website, just change my mind and leave. Other times I go to a website and I’m incredibly annoyed because it’s not working properly. So when, how, with Clarity, how do you work out whether a user is just leaving because they just changed their mind, or the user is leaving, because they’re incredibly annoyed?

RY: Yeah, that’s like the core purpose of why we built the product is to understand frustrated users and there’s multiple ways we kind of help you with that.

One is kind of looking at our customer experience metrics. For example, we have a metric called dead clicks or rage clicks, and just giving an explanation of those, dead clicks are where users are clicking on parts of the page expecting it to do something and this is quite common, where something maybe looks like a link, but isn’t a link, and they’re trying to click to learn more. Or there’s rage clicks, where they expect something to happen, but it doesn’t. So they’re click, they’re kind of clicking very aggressively in a short period of time. So you can probably you know, sympathise with this where you probably had a submit button that didn’t work, and you just went, you know, ham on clicking that button, trying to make it do something.

So those are really good indicators to understand when a customer is frustrated and even things like excessive scrolling, if they’re scrolling up and down a lot on your page, something is really confusing them. So those are kind of common metrics that we can easily capture and you can drill down to sessions, where those happened, and see what’s going on. And on top of that, you know, when you’re drilling down into these recordings, you can actually follow the journey, what did the user do from one page to the next page, and what potentially led them to getting stuck or frustrated.

And then a good other way to look at it is through aggregate data like heat maps, and heat maps are basically an aggregate data of all of the engagement that’s happening on the page, whether it’s clicks or scrolls and you can actually see areas where people tend to click more when it’s not supposed to be a place that we want them to click. So that could also be because they’re confused and frustrated, and they’re trying to do something there.

AV: You know, I’ve always loved the term rage click, it is just so evocative, isn’t that I’m sure we’ve all been there, it’s something so annoying, and you’re just clicking away. And I guess, the dead clicks too, I mean, I, again, I’m completely showing my age here but obviously, when I started out, a link was always obvious because a link always had a had to underline text, and it was completely clear what was a link and what isn’t but I think as the web has got a lot more advanced, it’s not as obvious now, what is the link and what isn’t. So I can absolutely see how having the ability to see what people are expecting to click on and what, what are actually links can be pretty important. I suppose particularly if you’re trying to do conversion rate optimization.

So what are the sorts of things that tend to break on a website?

RY: Good question. I think frankly, if it exists, it probably can break.

From my experience, we’ve seen the craziest things break. But maybe I can tell you is like an interesting story, first, of something that we actually noticed on Microsoft Bing.

One day, we saw on Reddit, a user posting that they were seeing their ads on Bing, in purple. So the background of the ad was in purple and of course, that’s not very good for engagement or conversion. So we investigated through clarity to understand, okay, why are some users seeing purple ads, that’s not something we want them to see and looking at some of those sessions, we noticed that all of these sessions had a little shield icon that’s green with a checkmark on each of the URLs, so they had some additional extension or software that was running on the programme that was adding these shield icons. And after investigating what that is turned out to be a popular antivirus and popup blocking and ad blocking software that, you know, if you had this installed, and a website was still showing ads, they basically messed with the CSS and changed all the ads to purple. Right. And we once we discovered that and made a fix, you know, then the ads started becoming white again, and you saw a lot more conversion and ad revenue going through.

So even though these small things that you don’t really expect, because hey, why would the CSS ever change for me? Do break. And that was one of those interesting things we were able to take away from with clarity.

AV: Yeah, I can see that. And you’re particularly I mean, obviously most of the listeners to this podcast, don’t have you know, a huge team of QA engineers who are who are going to go over their, their e-commerce websites. And you know, when we think about the enormous number of devices that that are now, you know, from mobile devices to tablets, and then obviously on desktops and laptops. There’s such an array of web browsers and even different versions of the same web browser. But to actually attempt to check the rendering of of your website across all of these platforms I would say, is essentially impossible, I think for for most people listening to this, unless, unless you’ve got the resources of some giant company behind you.

So I guess, putting the site out there, and then dealing with the problems through a product like clarity is the more practical option, even I guess, with, with us, you know, with, with our, our traffic, which, and again, for most of us, I’m sure most of us have less traffic than Bing does but even still, we’re going to end up with pretty large numbers of screen recordings.

I mean, how, how would you, how would we kind of work out, which are the ones that we should be looking at?

RY: Yeah, that’s a common challenge for anyone that’s trying to use a behavioural analytics tool and in fact, we know we do internally analysis and a lot of our Microsoft products, and like we came up with a small strategy that I can share, and that kind of helps anyone that’s trying to use a tool like Clarity to perform analysis in a more effective way. And usually, the first step in that is trying to determine what the purpose of my analysis is, right?

Usually, you have a metric or a KPI that you’re targeting to learn and improve, right? Whether it’s how do I improve my signup flow? Or, how do I improve conversions on these particular buttons that I have on my landing page? So starting with some of these goals in mind is a good place so you don’t kind of just get lost in the world of recordings. Then usually, what I started out with is a heat map, click heat map or a scroll heat map, which gives you the aggregate view to see where clicks are distributed or where the scrolling is happening and how far. For example, simple things like are users even seeing my call to actions or my key buttons that I need them to interact with? Then when I look at where they are interacting currently, then in Clarity, we have an option to go from a clicked region or interact region to a session recording to see where the users engaged with that region. So this really helps kind of look at situations where users led to clicking a particular thing, or led to clicking a particular piece of text, or visited a piece of a particular page.

And then you can look at the steps before and the steps after to understand what happened, and this usually helps a lot in discovering areas of improvement as well.

And then the biggest thing people tend to forget is like taking advantage of filters. We have tonnes of filters to help you slice and dice your data, like device type browser, even things like UTM parameters to help you filter down sessions where you came from an ad campaign versus a social media campaign. And all of these provide you that endless possibilities of looking at specific user sessions. And I would also look at comparisons, right, you can also compare heat map A versus heat map B through clarity and you can understand why there’s a particular change that’s happening on my website, compared to what it was maybe a couple days ago.

AV: Oh, I see. So filters, as well as sort of across devices, it works over dates and times as well?

RY: Exactly. And then as I mentioned earlier, some of these customers experience metrics like rage clicks, and dead clicks are also good ways to jump into problematic areas very quickly. And you might see multiple types of problems that all kind of compressed, combined, into rage clicks and then you can dissect them.

And then I think once you find a problematic session, it’s very easy to use the session metadata to find other sessions that are similar to that. So maybe a particular device type caused this issue, so now you can filter to that device type, and look at other recordings to see what happened to them.

And if you’re not doing any of this analysis, you know, you don’t you don’t know what metric or KPI you’re trying to target. Well, that’s a challenge in itself, but the general thing I suggest to most people that you know, on a website is watch 10 to 20 recordings a day, just feel for your product, feel for your customers, and just, you know, step into their digital shoes and see how they’re interacting with your page.

AV: Yeah, no, I see that’s really powerful. I mean, I think I mean, to be honest, I think most of us probably rely on customers actually just making a complaint that you know, customers are struggling with something and then send an email or or go in the live chat and make a complaint. But I always think for every one customer that makes a complaint because something’s gone wrong, there’s probably 100, who’ve had the same problem, but aren’t going to go to the effort of reaching out to tell us what the problem is. So I can see that this is going to be really helpful.

So how do we find out if a particular page is confusing users?

RY: So that’s an interesting problem and it’s kind of hard to detect if you don’t have this behavioural data. For example, sometimes simple things as scanning and reading pattern of how users traverse the page is actually very helpful.

For example, one of our customers, were that we are going to be publishing a case study with soon, mentioned that they deal a lot with legal jargon on their website and as users actually read that content, they’re actually using their mouse to move around on the text and they can see that sometimes they go back and forth, which basically is indicating that that legal jargon might be too confusing or kind of too hard to comprehend for an average user. So they can kind of improve that in their next iteration.

Or even things like excessive scrolling, as I mentioned before, or even the time spent on the page can be an indicator, like why are they spending so much time on this page, it’s a very short page or small amount of content and that could be good indicators of what might be confusing users.

AV: And I read the Imagely had a success on their conversion rate using clarity.

RY: Yeah, Imagely is a build WordPress themes and plugins for the photography industry and they, they’ve basically found an opportunity with Clarity. So they looked at their pricing page, which is you can imagine, it’s kind of like a table and they had different headers and different features and different pricing options for those. And what they realised was looking through session recordings was that users were like clicking on the headers of the different features, and it wasn’t doing anything. As you mentioned, people assume things are clicks, and nowadays, because of just how different user experiences, and then the main reason they found out is through session recordings and they made a simple change, which was to implement a hover tool tip, right? When a user hovers over those section headers, it just shows a pop up explaining what that feature is and that it led to increasing their conversions by 25%, which is quite great. And it’s a very small problem, but it led to a good opportunity for them.

AV: No, that’s great. And it’s hard to see how else they would have found that out.

And then another story I heard was Travel Boom, could you tell us about them?

RY: Yeah, Travel Boom provides an online marketing services. They have a customer that they manage called Myrtle Beach Golf Trips, and their goal was to basically help improve the conversion rate for their customer. And they had kind of key, key call to actions, shopping and requesting a quote. And the first thing they did was open up our scroll heat map and they quickly saw that, actually, an average user doesn’t even see most of their call to action, actually, less than 50% of their users actually see those buttons, and this was across both mobile and desktop.

And basically, it’s because it was below the fold. And for folks that don’t know that expression, you know, above the fold is what we refer to as the content that a user sees when the page first loads, and it’s above where they have to scroll. So they basically saw these call to actions below the fold. So it was a very quick insight. And they did couple A/B experimentations to move some of their call to actions up higher on the page. And it led to a 30% increase in conversions. And I can’t emphasise this enough of how much people don’t realise what content is above versus below the fold. And that really drives a lot of engagement by customers.

AV: And I think as well, because often the people that develop pages have quite big screens and the people viewing the pages tend to actually have much smaller screens, I guess is probably part of that.

RY: Yep, exactly. So and if you want to learn more about these case studies and other case studies, we do have them published on our site at

AV: That’s great.

So you sold me on the product, it sounds, it sounds really valuable. How much does it cost?

RY: Completely free.

So Microsoft clarity is a completely free product. We really want to empower everyone to build better products through behavioural analytics and we really wanted to democratise this product because we saw a huge benefit for us when we use it internally. So we wanted to provide that goodness to everybody

AV: And great.

And are there a limited number of impressions or a package you need to upgrade to?

RY: Nope, there’s no upgrades, there’s no upsells, you don’t have to, there’s never going to be a paid tier. Our sessions, we capture your full traffic, we’re not sampling or anything. So you can experience the best of of everything and get all of the data you need to make good decisions about how to improve your website for your users.

AV: Well, that is absolutely fantastic. And I guess if you see any benefit from it, you’ve technically had an infinite return on your investment, if it was free.

And then in terms of installing what, is it just the usual, they put some tags on on the website, I guess?

RY: Yeah, it’s very similar to any other tag installation, we have a piece of JavaScript code that you add to your head section. And, if you’re using things like Google Tag Manager, you know, you can install it in less than two seconds, right? You can find ourselves there as well. And if you’re using other plugins or other platforms like WordPress, Shopify, Wix, we have detailed instructions as well and how to integrate.

AV: Well that’s absolutely fantastic.

Thank you very much for your time and for telling us more about Clarity.

RY: Thank you so much for having me, Andrew.

AV: So if you have a Shopify store or other e-commerce platform, I’d strongly recommend you install Microsoft clarity.

And also, of course, Machine Labs.

We’ll help you really understand your customers by downloading a customer database, studying all the order patterns and behaviours and then looking at 950 other demographic variables. These are things like the customer’s age, their gender, their income, whether they’re in a rural area or an urban area. Maybe whether they have a dog or a cat, whether they have a partner, whether they’re, they have a family, whether they have a garden, what type of car they drive, and if the customer is in the United States, we’ll also look at whether they’re a Republican or a Democrat. We then put all of these factors together to make personal product recommendations, and personalised emails so that you can sell more.

So see you next time on the joy of marketing

Reader Interactions

Leave a Reply

Your email address will not be published.