Host: Hey everyone, what’s up? And we are back after that amazing session from Shane. I think it was very insightful, and we’re back for another session. But before we begin, I would like to thank all our sponsors, all our partners who have partnered up with us to make this event possible. I would also like to thank all our guest speakers who have been with us up till now and who will be with us in the next few sessions as well.

I would like you all to sort of show some love to our speakers and the partners using the emojis. Let us know in the comment section how you’re liking the event. If you have any feedback for us, please let us know, because obviously we want to become better. We want to do better events for you guys. So make sure that you drop your feedback in the comment section for us as well.

Now we have a very interesting session for you all. It’s called why good enough performance is no longer enough, and it will be presented by Georgi Petrov, who I would like to call up on stage. Now you might have heard that name in the past associated with different performance tools, with NitroPack maybe, with other performance tools as well. But as soon as Georgi is on stage I would like himself to introduce himself to the audience.

So if I can have Georgi on stage real quick. Yeah, Georgi. Georgi is here. Georgi, if you can really quickly introduce yourself and let us know what your background is and what do you do right now? What do you do to keep yourself busy? And then after that you can just start off with the session without any delay.

Georgi Petrov: Thank you so much. Thanks for the invite and super happy to be here. So basically I’m involved in the performance industry since 2013 when we started the first version of NitroPack. It ultimately grew into a popular, initially SaaS, then WordPress focused solution, and now serving a big chunk of the active internet. And currently I’m involved with a company called Uxify, which I co-founded together with my partner Mo, and we’re focused on user level optimization, monitoring, and AI on top of them.

Host: Yeah, thank you for that intro, Georgi. And for everyone else, Uxify is also our partner for this event. So thank you so much, Georgi and Uxify, for being a partner for this event. So Georgi, your session is about, you know, performance is not enough, good enough performance is no longer enough. Can you tell us a bit about what that session is and what will be discussed in this session?

Georgi Petrov: Right. So the core of the session is thinking of performance through the very practical meaning of why performance matters, and then identifying what is the good performance for you, and why actually it’s much smarter if you have a website which has a business function, like it can be e-commerce, can be SaaS, can be even hosting. Like why, when we have this website, you should look at performance from a growth perspective, from marketing perspective rather than from technical perspective, and how to correlate that connected to the business outcomes.

And this basically answers the question why good is not good enough. Because most of the times what we’re seeing with our customers is like, we’re seeing how actually they’re striving to go some crazy numbers of low LCP because they’re being out competed, and they see how their users are reacting on this LCP.

Host: Yeah, that sounds amazing, Georgi. And I won’t let our audience wait more. So the floor is yours, Georgi. You can start off with your presentation.

Georgi Petrov: All right. Share my screen. All right. So that’s my deck. Okay. Let’s start with a little bit of history and actually understanding why we are here today, in 2026, discussing still performance when actually it has always been the topic, one of the big topics in web.

So I will just briefly rewind till 2018. For the members who were doing performance at that time, please share, identify yourself and share a funny story. I can just share that at 2018 there was no Core Web Vitals, there was no LCP, and there was no anything that actually mattered for real understanding the real user, that there was nothing out there.

And in 2018 there was this saying propagated, like, basically there was a common industry knowledge where tools like Pingdom, like Kissmetrics, or like a few case studies here and there saying that if your website is 3 seconds, like, or lower, that’s super fast website. Basically 3 seconds was like the gold standard where everyone was striving for. That was at the time where basically the user patience was kind of, at that time was like the lowest it could have ever been, but still it was cons, like from the perspective of today, was like a very high.

And at that time also there was no, like, a singular, like, uniform way to measure performance. And like some people were doing page load together with all the JavaScripts and everything that is loaded together, page load time, other people were doing content loaded, and so it was very, very basically a dispersed effort across different people. And that’s why also generally the understanding of performance was not so high among website owners.

And then luckily in 2020 we had the launch of Core Web Vitals, which I think a year later they went into their Page Experience, and that solved a couple of issues. So the first one is a new standard appeared which was an LCP, and there was like the metric of largest paint. I’ll have it on the next slide to explain to the new folks.

But basically we introduced an LCP, where this LCP was considered to be the most reliable way to measure the perceived loading for any website. And why perceived loading is important? Because it’s designed to be for the users, right? So when we optimize, we want to optimize for the users. And perceived loading was where everyone was striving for in this initiative. And 2.5 seconds was chosen, but another interesting thing that happened was, like, it was not an average, it was not a median, it was P75, which means that it basically measures the 75% of the fastest user page views.

And the next, what happened after this launch is just history. Like it completely changed the performance industry, and then basically everybody started to understand the significance. Google highly propagated it through their efforts and huge performance efforts across the industry. Still they’re still ongoing. And yeah, moving up to 2026, we still use heavily these thresholds and the standards, which is great.

Right, yeah, that’s right, GTmetrix. I’m seeing just some of the comments and the questions, like, yeah, you can run GTmetrix directly. The thing is that, so basically GTmetrix was also popular, and of course this was like even one of the best metrics tools, because it was like very, very powerful and running, like, good so-called synthetic tests with today’s terms. And yeah, even, if you, people who remember from that time, there was a popular technique, because I heard the previous topic, one of the popular techniques was actually putting your images in sprites, right? Creating one big image and then moving it with CSS.

All right. So going back to today, today we have a very interesting opportunity and situation to actually start thinking about performance from a completely new perspective, and this is what we have been doing in Uxify recently. Like basically our whole value propositions is around the user, focusing about how the user reacts on performance and other type of frictions. Because, right, performance is by definition, like the way it’s organized currently, is by definition negative metrics, but there are also other negative metrics which are also real user metrics, and they are not directly performance related, but there are still frictions.

So moving to 2026, we have still this general good threshold by Core Web Vitals of 2.5 seconds. But what changed in the last six years? So one of the things that changed is that more and more websites started to get under the 2.5 seconds threshold, which is great, amazing. But also something else changed, and I will share with you what changed in the next slide.

But let me tell you, for the folks who are just joining, what is LCP? LCP is currently the most popular page load performance metric, which is in use by website owners, agencies, tools, like basically everyone uses LCP to measure how fast a page loads. And in very simple terms, LCP directly calculates the time and shows you the time of load of the largest element on the page. This could be the hero image, the headline, or like anything that is large enough to be significant enough on the page, that can be considered an LCP.

And this to some extent has, like in our own studies, it has about almost 80% of direct correlation to perceived load. Still there are 20 something % of LCP which does not directly correlate to perceived load. So think of, like, pages where they have multiple elements, like think of a grid where every image is the same, and it’s like a big grid, very often it will select the first element. So LCP has its own problem, but in 80% of the times it’s correct, right, to understand the perceived load.

And if it is correct to understand the perceived load, which means that we can measure the user patience with it. But now there is a new challenge, which is the following. User patience is different across industries, across websites. It’s different sometimes even within the same website across time.

So depending of how you market the website and who actually is entering your website, from what type of device, what is their previous experience. Like one of the big contributors of changing of the user perception of how fast the website is, is actually how fast is also your internet connection. Because if you’re a user browsing for a very fast internet connection, you tend to expect every website to be fast, right?

And that sees everything piling up into different cases where fast website actually means different for every website, for every audience. But, right, still everyone, in every industry, in every website, will still use 2.5, right? And that’s a very good start, but I will share a bit more. What’s the patience of your users is the biggest question here, and how to find it out.

So in the next slide, basically I will walk you through a decision matrix which we use for our product and for our customers, which we have proven to be good enough to find opportunities for increasing engagement, increasing revenue, and finding where exactly performance improvements matter.

So before starting, we need to measure performance, right? And there is very often this case of, especially from the smaller sites and just joining, and agents managing small sites, like very often a Lighthouse, which, like you can take it in variations, it can be within, it’s free within Google DevTools, and it can be multiple tools on the internet that do the so-called synthetic test, which is basically Lighthouse, which is amazing. I mean, nothing bad about it.

But it has one main use case. First is to be used on a staging website, and the second main use case, which probably secondary, is to use, still you can use it on a production website, but only for the purpose of finding different deviations, like anomalies and changes.

Now the big point is, like if you have a live website with real users, you want to look at your real user performance, right? So it’s basically the same LCP, but the way that is being experienced by the real users. And the rest of my topic will focus on the real user experience. So advice here is, you have a staging website, go with Lighthouse. You have a live website, go with real user monitoring tool. That’s it. That’s the advice we go with.

Okay, the first thing with decision matrix, like how fast I should go with my LCP, like what is a good LCP for me. At the very start you start with a 2.5 seconds baseline, right? That’s the good enough performance which Google suggests. If you’re passing it, you basically move into the passing Core Web Vitals, given that the other Core Web Vitals are fine as well. This is the floor. Like this is where you start with.

If you’re on more than 2.5 seconds, that’s easy, you have work to do. I mean it’s easy to understand where you need to be heading, you have work to do, get under 2.5 seconds. This is the first requirement that you need to cover in terms of LCP.

Now it gets more interesting. Of course, this is the, like you can use many tools, but like using the CrUX data. This is the Google PageSpeed tool, where basically the first part of it is the important part, where it shows you what is the LCP and INP and CLS of your real users as measured from Google Chrome data. Right. Okay.

So how fast I should go with LCP? Number two, go and find out the industry average LCP. Like you’re in e-commerce, fine. You’re in the US selling in the US, great. You’re selling, like, health products, fine, great. You are in this industry, you’re competing with it, go and find this industry average LCP. Like there are tools, the real user monitoring tools who offer this information already, where also have this information. You go and find, like if 2.5 seconds is for every website, including the ones that are, like, not really relevant, go and basically understand what is your industry LCP. And it’s a game of competition.

So knowing where you stand, and what is the expectation for your industry, is the first thing we need to know after the 2.5 seconds. More and more industries which are more competitive are also very competitive on LCP. And if you’re in a very competitive industry, the chances are that your industry average LCP is very much lower than 2.5 seconds. Yeah, right.

That’s an example of what a comparison to the industry would look like. It’s from our tool. You can get it from any tool, of course. And that’s on the industry.

Now we move down the funnel, and now we look at specific competitors. So specific competitors are the ones, like not in general all competitors, not necessarily the one that you share some customers with, although it’s still super relevant, but the most relevant group of competitors are the one who you share traffic with, right? So this will be the ones that people will go and click, like on search or on like LLM search, they will find you, they will click a few other people, and they will start browsing with you. So these are the ones who are sharing traffic with, and they’re the most relevant competitors.

Why it’s important to look generally in competitors and in industry is because your users aren’t comparing you to the Google thresholds. They’re comparing it to the last site they visited, and that’s usually, if they’re researching now, they are usually a site of a competitor or another site that might be faster enough. So you should be focused on that side. That’s a tip.

Like one example how a competitor comparison might look like in your favorite RUM tool. And you can compare how you stand against them, right?

And now we get to the interesting part. That’s finding your user LCP sensitivity. So that’s something that is measurable but is not super straightforward with the current tooling. Right? Why it’s not super straightforward? Because it involves multiple data points, and it involves platform specific knowledge, where most of the tooling are platform independent and platform agnostic. And in these cases, like there are already some correlations available on the market, but in these cases, like that’s the most important way to look at LCP, if you look at it from practical standpoint, from the business standpoint of what is a good LCP and how fast it should get.

And I’ll share with you a few examples from a customer of ours I just pulled before the presentation, and show you how this user sensitivity on LCP on your website actually makes the needle, and how it completely switches the question, moves the needle. Yeah.

So first of, let’s start with engagement and conversion. So when you start looking at your own real user data and how your users react, you will start seeing graphs like this. Like you can look how the page views per session are basically improving with the LCP, right? That’s first engagement signal, what should be your LCP. Second engagement, like you can look at how the LCP by pre-conversion, for in this case by add to cart, how basically you get add to carts, more add to carts with the better, right?

And then you have, see how basically you get more checkouts with the improved. Similar correlation here, like checkout rate upside. That’s you see how checkout rate is different at different LCP levels. Like we are seeing here a data of, like I guess I think about 2 million users, and that’s a very good upside of what actually is achievable as a correlation of different LCP. And when you start monitoring these type of situations, you tend to understand the potential of looking and improving your LCP, right?

And one bonus benchmark which you can look, and we also following very closely with different frictions and, like, performance metrics, is how you are signaling different engagement, like for example how you’re signaling dwell time for your purposes with LCP. And here you see the distribution of LCP again by second and dwell time and the opportunity size there. And here also you see how your competitive landscape look like in terms of LCP and who you should be outranking.

And these are all suggested, like something I highly suggest for you to follow if you’re optimizing for the bottom line your website, and if you want to use the LCP and other performance metrics as growth levers compared to just for technical, like quality of life optimization. And ultimately you should end up with something that is target LCP, and that’s where your target is. So basically that’s the decision matrix of how to think of LCP from the practical standpoint of improving your bottom line.

But, and yeah, ultimately your goal is to find the ideal LCP. Once you find it, it’s very often much lower. Okay, sometimes less lower, but it’s definitely lower than 2.5 seconds, the suggested ones. And that’s why how fast is fast enough, and why good performance is not good enough nowadays, is a real question for decision making and for businesses, right?

And I’ll give you a brief also matrix, which is rather suggestions. Say, we’re giving you also something we also have as an optimization tool. So how do I get to my ideal LCP? So very often, like mid-market brands and upper SMBs, they will have someone in house who is in charge of performance. Might not necessarily be a dedicated role, but it could be like in the web team or marketing team, who is, one of the tasks he has is to deal with the performance.

So usually this type of person is the first go-to person to actually go and try to talk and see what we can do about performance, once we see the real user sensitivity and the competitors and the industry benchmarking. This data-driven approach is useful for performance engineer or the person tasked to know where exactly and how exactly to look. At the most of, like in Uxify for example, you have a drill down with 100 plus segments where you can drill down and actually identify the groups, and have the AI agent on top where you basically can find where exactly is the issue with the LCP. And, like that’s the go-to strategy.

This engineer sometimes might use, like, optimization tooling, and like for example like Navigation AI of NitroPack, our previous components, like amazing optimization tooling for WordPress, pro another optimization tooling for WordPress. Cloudflare have bunch of rather generic type but still very reliable optimizations. And yeah, that’s all piles up together. Still it’s a good strategy to use someone who is aware of what needs to be done, like rather than going in applying multiple tools at once. It’s really gives the best results.

And yeah, nowadays we have a lot of AI agents who tend to more and more understand the code. We for example use a lot of AI in order to propose some meaningful fixes and to find hard to find situations where it would require a performance engineer. Some of his tasks are actually taken by this AI agent, identifying weak points in performance based on the real user data.

And for those of you who are just hearing the real user data and situation, and wondering why do I need that when I just can run a test and tell me the score. Well, the difference is big. If you’re running a live website, it’s a huge difference. And the main difference is that it’s not only the website itself, the moving piece of the puzzle. There’s a lot of moving pieces, with network conditions, with VPN restrictions, like with device power, internets on and off for the users, like going hopping on and off different locations. And all this is actually resulting in the real user experience, knowingly or unknowingly by the user.

And you can do more, like for example there’s already certain, like very strong motions going in the user level optimization phase where we are involved, and basically keeping the user closer to you regarding on the environment. And yeah, that makes a big difference to actually optimize for the real user.

And that basically, focusing the real user and focus on the business first perspective of actually what does it mean for my engagement, for my user, for my website, they’re all aligned. We ultimately, with this focus, we ultimately achieve the same thing, and it is making your user happy, right? Because if you’re performing, if he’s happy, he, she is happy, she will continue this journey. The button that is clicked immediately responds visually, and then she clicks, she tends to click more, and then she tends to return at a higher chance rate to your website, because she remembers that website was very, like subconsciously remembers it just worked.

And this all piles up together, and it basically fundamentally results into higher engagement and more conversions, and it fundamentally change the ownership of the performance task, right? Because, like if from landing, the performance helps the users, like Uxify, like the journey is how we call it, or basically improve the user experience of the journey. And this correlates directly to success, more successful journeys. That’s now a business topic, right? That’s now a business question, that’s a management question, that’s, like how we are more successful with that.

And that might be on the marketing, that might be on growth, or any team that is tasked to make this website work. When it works for the user, it works for the business as well. So basically we’re changing the question from technical question to business question, and when correlating to engagement, conversions, SEO, right? That’s all, folks. Thanks so much. If you have any questions, let, hope it was interesting.

Host: Yeah, thank you so much, Georgi. This was fantastic. Very insightful. We have some questions in the chat as well that I would like you to answer. We have a few minutes, and we can use that to answer questions from the chat. So Veronica has a question, and she asks that, how can we actually measure how LCP influences my revenue. So I will pin that on the screen so you can read it. It should be on the screen now. Can you see it?

Georgi Petrov: No, I don’t see it.

Host: All right, let me just read it out. So it says, “How can we actually measure how LCP influences my revenue?”

Georgi Petrov: All right. Yeah, so it’s just basically using the real user data. And this real user data already has the answer, right? So we already have the real user data, you just need to visualize it. And that’s, you use the real user monitoring tool. That’s the easiest way. Just the real user monitoring tool should support this visualization.

Host: Awesome. I will now go to the Q&A section, and we have a few questions over there as well. Let me scroll down to see if we have any questions for your session specifically. I saw a few questions around LCP and I wanted you to take those. So Hamam has a question. How can design systems, no, so sorry, Yashi has a question. How do modern web technologies help developers exceed good enough performance standards?

Georgi Petrov: Yeah. So there is a, how to say, modern technologies do not necessarily mean faster LCP, right? Because the motivation number one of every modern technology is to be easy to build, which is slightly different motivation than performance for the visitor. And it’s been like that for quite some time, with all the frameworks, and now with all the vibe coding and things like this that is going on, and all the technologies that you pile up together.

So when you pile them up together, you definitely optimize for your productivity to build this website. But performance is rather an optional add-on to most of these technologies rather than something that goes with them. So we’ve seen some modern, like for example teams that, like we have one or two teams which they have put some constraints on their buildability, on their technology, but they have really produced a performance first team. We’ve seen that. But generally the modern technologies in general, then they, I don’t know, have performance as first priority in reality. Although their marketing might be saying something different, it’s often not the case. Like performance usually goes in, does not go together with easy to build, right?

Host: All right. Makes a lot of sense, Georgi. We’ll take the last question for your session, and then we will just wrap it up. The question is, Georgi, could the pressure to exceed good enough performance lead to unsustainable expectations for workers and organizations, and how should this be managed?

Georgi Petrov: Yeah. So the first part of the question is, it might really be a bit tough. But it’s usually, like you need to understand why we’re doing it and what do you need. So from there it simplifies a lot of situations. So if the teams who are building it, they consciously know that they need to strive for that, they willing to take tough decisions. Like tough decisions like, we don’t want that big video here, we’ll do something else, right? We don’t want that big JavaScript which is here, we do something else.

And they, in this way, when they think technology first, like from performance first perspective, they’re able to take the right decisions from the beginning of what is possible and what is not possible. Because too many of the nice things that they want to do are actually not beneficial for performance. Yeah.

Host: All right. Thank you so much, Georgi. This was it from this session. Folks, I would like to thank Georgi for being here and presenting very insightful presentation and a demo. I’m sure a lot of people will take a lot from Georgi in this session around LCP and, you know, why it matters so much to your website and to your client’s website based on what area of work you are in. So thank you so much, Georgi. This was it from this session. As for you folks, don’t go anywhere, because right after this we have a rapid fire quiz. So if you missed out on the last activity, this might be your chance of redemption. And you might end up winning some amazing gift cards. So don’t go anywhere.