Core Web Vitals in e-commerce and widgets - how not to slow down your online store?

A widget that increases conversion but slows down the store might end up costing you more than it earns

Do you know this scenario? Customers very often come to me with one burning dilemma. They say it straight: “Listen, I want to increase conversion with popups and widgets, I care about saving abandoned carts, but on the other hand, I am panically afraid that I will kill my store’s performance.”

And you know what? They are absolutely right to be afraid. As an e-commerce optimization expert, I perfectly understand this temptation. Most stores want to have more popups, information bars (sticky bars), dynamic recommendations, or social proof notifications. We see these mechanisms at the biggest players and believe that this is a magic “sell more” button.

Unfortunately, rarely anyone realizes the hidden technical costs. The truth is brutal: every additional script plugged into the code can drastically slow down the site. When you attach subsequent tools without control, you immediately worsen Core Web Vitals. As a result, Google cuts your visibility, and frustrated users leave the store. Let’s look at the hard data; reports such as the Portent study clearly show that a site that loads in 1 second achieves up to a three times higher conversion rate than a store that needs 5 seconds for that.

When I conduct an SEO audit, a certain pattern repeats almost always. The marketing department invests in great tools, but LCP (Largest Contentful Paint) optimization completely fails. Why? Because an external, heavy chat or popup widget loads first, blocking the loading of the main product photo. The user sees a white screen.

That is why in my work I always educate clients: the biggest problem today is not whether to use widgets at all. Of course, it is worth it because they are powerful marketing levers! The challenge is how to use them wisely and lightly, so that page load speed does not become a victim of your marketing.

We must stop treating performance as a programmers’ problem. A fast online store is not exclusively a technical detail; it is a full-fledged, real sales factor. Even the most beautifully designed discount pop-up will not save your business if the customer manages to close the browser tab before it even manages to load.

What are Core Web Vitals and why do they matter for an online store?

When I talk to business owners, I often hear: “After all, this small bar at the top of the page weighs only a few kilobytes, how can it spoil my results?”. The problem is that Google evaluates site performance very rigorously. According to the Web Almanac 2022 report (published annually by the HTTP Archive), external scripts (so-called third-party scripts), which include marketing widgets, are the main cause of problems with CLS and INP and delay the rendering of pages.

Let’s see how these small additions hit individual parameters.

LCP - the biggest e-commerce problem

LCP optimization is an absolute priority today. LCP measures the time in which the largest, most important element appears on your client’s screen (e.g., a beautiful promotion banner on the home page or the main photo of a shoe on the product card).

If a widget (e.g., a chat script) loads too early and does not have a loading priority set, it starts fighting for bandwidth with your content. The browser, instead of showing the customer the product, first processes the marketing code.

Without bluffing: I will say it straight, based on dozens of audits: in many stores, it is not oversized photos that kill Core Web Vitals LCP. Heavy, poorly implemented marketing scripts kill it.

INP - when the store starts thinking for too long

INP (Interaction to Next Paint) is an indicator that tests the responsiveness of websites. It checks what happens when a user interacts with the site. Imagine that a customer is browsing dresses on a phone and clicks on a red color variant. If in the background the browser is burdened with grinding a powerful JavaScript file from a cart-saving popup, the page will simply “freeze” for a fraction of a second.

Interactivity suffers most when JS is too heavy. The user scrolls, clicks in the menu, or selects variants, and the page reacts with a delay. For the customer, this is a clear signal: this store is broken.

CLS - when something suddenly jumps on the page

Cumulative Layout Shift (CLS) is the bane of mobile devices. Has it ever happened to you that you wanted to click the “Add to cart” button, but in the same millisecond a sticky bar with a countdown timer slid down from the top, the page jumped down, and you clicked on a link to the regulations? That is exactly what fatal CLS is.

Poorly loaded popups, information bars, and floating social proof notifications move the page layout. Disturbed interface stability is a simple path to customer frustration and an immediate exit from the site.

How widgets affect LCP, INP, and CLS?

When I analyze my clients’ e-commerce, I see a repeating pattern. Entrepreneurs invest powerful budgets in attracting traffic, but forget that a battle for fractions of seconds is going on behind the scenes. To understand where you are burning money, we must look under the hood and discuss the CLS, INP, and the famous LCP indicators.

As results from the annual, extensive Web Almanac report (2023 edition) published by the HTTP Archive show, external scripts, including among others popup and chat systems, are one of the main culprits lowering site performance and spoiling UX. Let’s see exactly how they destroy your results.

LCP - the biggest e-commerce problem

For every store owner, LCP (Largest Contentful Paint) optimization is the number one task today. This parameter evaluates the rendering time of the largest element on the screen. Usually, it is a large hero banner (on the home page) or the main product photo (on the product card).

The problem is that very heavy widget scripts drastically delay the loading of this main element. If a marketing tool loads too early (synchronously), it starts brutally competing for connection and processor resources. Your customer’s browser, instead of downloading the photo of the shoes they want to buy, is busy processing the code from the pop-up window.

In many stores, it is not photos that kill Core Web Vitals LCP. Marketing scripts kill it. This is a strong thesis, but I sign it with both hands after conducting dozens of audits.

INP - when the store starts thinking for too long

From March 2024, Google established INP (Interaction to Next Paint) as the official standard for measuring responsiveness (replacing the outdated FID indicator). INP tests a simple thing: how quickly your store reacts to a click.

Site interactivity suffers terribly when JavaScript code coming from external tools is too heavy. Imagine that a user enters a product card and clicks on a different size variant. The widget’s JS file burdens the browser’s main thread, so the page “freezes” for a fraction of a second and highlights the selected size with a delay. Such sluggish widgets can worsen responsiveness during various actions on your website: scrolling (e.g., while viewing reviews), clicks in the navigation menu, selection of variants and adding to the cart. A store that “thinks for too long,” lags, and gives the impression of being broken, immediately lowers trust in the seller.

CLS - when something suddenly jumps on the page

Cumulative Layout Shift (CLS) measures visual stability. This is probably the most irritating problem for users, which has a gigantic impact on conversion.

Recall how you felt when you wanted to click the “I’m buying” button, your finger was already heading toward the phone screen, and in the same millisecond a sticky bar with free delivery loaded at the top. The entire page layout suddenly slides down, and you, instead of the cart, click on a link to the return policy. Popups, info bars, and social proof widgets (“Kasia from Wroclaw just bought this”) have a terrible tendency to move the layout.

Poorly loaded elements affect interface stability and are a real problem especially on mobile devices, where every pixel of the screen counts. If you have a high CLS, your customer simply has no chance for convenient shopping.

Most common store mistakes when implementing widgets

When I open the code of a new e-commerce site to conduct a store’s SEO audit, I usually see where the customer’s money is escaping after just a few minutes. The intentions of business owners are fantastic; they want to engage buyers and increase sales. Unfortunately, the technical execution of these ideas often generates critical Core Web Vitals errors.

To build and maintain a fast online store, you must know what to avoid. Here is a summary of the most popular capital sins that respond to very specific performance problems.

Loading everything at once

This is a mistake I see in 8 out of 10 audited stores. The owner installs a chat widget, a popup for an abandoned cart, and a promotional bar, after which all these scripts load in exactly the same millisecond in which the user opens the page.

For a browser, this is a murder. Instead of focusing on rendering the product photo and price, it must process the code of external marketing tools. A user does not need to see a newsletter popup in the zero second of the visit. These elements should be loaded asynchronously, in the background, with an appropriate delay.

Lack of lazy load and conditional rendering

Lazy loading is mainly associated with photos, but scripts should also be subject to it. If your widget with recommendations or customer reviews (social proof) is located at the very bottom of the product card, why do you force the browser to load it right at the beginning?

The lack of conditional rendering is a massive waste of resources. Scripts should be called only when the user approaches a given section (scroll) or performs a specific action.

Several tools doing the same thing

This is a classic example of technical mess, which we often call “the failure of abundance.” The marketing team adds tool A for exit-intent, tool B for collecting leads on mobile, and tool C for displaying a sticky bar.

Each of these tools pulls its own, often the same, code libraries (e.g., jQuery) and sends separate requests to different servers. The browser must perform triple the work. Consolidation of tools into one, comprehensive solution is the fastest way to recover lost seconds.

External widgets without performance control

Pasting a small fragment of code (<script>) provided by an external provider seems safe and fast. However, we forget one thing: by pasting this code, we give away part of the control over our store to a foreign server.

If the infrastructure of the widget provider has a failure, is overloaded, or simply poorly optimized, your own store will suffer. According to analyses published in annual Web Almanac reports, it is third-party scripts that are one of the biggest performance blockers on the web.

Lack of regular SEO and performance audit

Performance in e-commerce is not a state given once and for all. It is a process. You implement new features, change banners, update the store platform, and add new tracking ad pixels.

Even if your store was a speed demon on the day of release, after half a year of intense marketing activities, it may already be “getting short of breath.” The lack of regular monitoring of indicators makes you the last to find out about performance problems - usually at the moment when you look worriedly at a drastic drop in revenues in Google Analytics.

How to add widgets and not slow down the store? 7 practical rules

When I explain to clients what a destructive influence poorly implemented scripts can have, I often see terror in their eyes. They wonder if they must now remove all tools to maintain a fast online store. The answer is: absolutely not!

The art lies in wise resource management. As a practitioner, I have gathered 7 specific steps for you that show how to improve Core Web Vitals without giving up marketing activities. This widget optimization is your key to higher conversions.

1. Load widgets only where they are really needed

This is the absolute foundation. If you have a popup with a discount code that activates when trying to leave the cart (exit-intent), why does its script load already on the home page or on the blog? Configure the tools so that the code is loaded only on those subpages on which it is actually to appear. Fewer requests mean faster loading.

2. Delay the loading of non-critical elements

The principle is simple: first we load what the customer came for (product photo, price, “Add to cart” button), and only then everything else. Use defer or async attributes for marketing scripts. You can also delay the call of the chat widget by 3-5 seconds or make it dependent on the user’s first interaction (e.g., when they start scrolling the page). Thanks to this, the script loads in the background, and the user immediately sees the offer.

3. Avoid heavy libraries and redundant JavaScript

I often find that a small widget pulls behind it huge, outdated libraries (e.g., jQuery), while modern browsers handle lightweight, clean code (Vanilla JS) very well. Choose software providers who care about the lightness of their solutions and use modern programming standards.

4. Care for layout stability

How to avoid a CLS indicator shining in red? Reserve space for dynamic elements! If you know that a sticky bar with information about free delivery will appear at the top of the page, let there be an empty space reserved for it in the CSS code immediately. Thanks to this, when the bar loads, the rest of the page will not suddenly “jump” down under the customer’s finger.

5. Test mobile separately from desktop

Remember that your powerful office computer with fast fiber optics will process every widget in a fraction of a second. However, your customer is on a train and browsing the store on a few-year-old smartphone with poor range. Always evaluate performance from a mobile perspective. In e-commerce, it is smartphones that generate the majority of traffic today, and Google evaluates your site specifically through the mobile prism (Mobile-First Indexing).

6. Do not measure only PageSpeed - look at real user data

The Google PageSpeed Insights tool (laboratory data) is a great signpost, but the most important are the so-called Field Data, i.e., data collected from real users (Chrome UX Report - CrUX). They decide how Google will evaluate your Core Web Vitals. Focus on how the site is felt by living people, not just bots testing the server from one, ideal point.

7. Regularly do an SEO and performance audit

Implementing widgets is not a one-time action. When you install a new plugin, change the template, or add a new tracking pixel, performance can drop. A periodic technical review allows you to catch “parasitic” scripts before they start noticeably lowering the conversion.

How to recognize that it’s widgets spoiling your SEO results and conversion?

Very often, business owners live in ignorance. The store looks nice, widgets display correctly, and yet something grinds in the financial results. Before we do an advanced SEO audit, it’s worth looking at the business symptoms. How to recognize that page load speed is a victim of your marketing?

Visibility drops despite good content

Do you invest in great product descriptions, run a blog, build links, and your positions in Google systematically drop or stand still? Since the algorithm updates in 2021, Google ruthlessly verifies UX. If Core Web Vitals are down due to sluggish widgets, even the best text will not position your assortment in the TOP3.

Worse mobile than desktop

Look in Google Analytics at the conversion rate divided by devices. Of course, mobile conversion is usually slightly lower than on desktop, but if the difference is drastic (e.g., 3% on computers and 0.4% on phones), it’s an alarm signal. Smartphones are extremely sensitive to being “overloaded” with scripts, which immediately kill mobile performance.

High bounce rate on landings

You launch a new campaign on Facebook with a very precisely selected target group. Click-through rate (CTR) for the ad is great, but in the store’s statistics you see a gigantic bounce rate. Why? Because the user clicked, waited 4 seconds for a white screen (due to delayed LCP), gave up, and went back to scrolling Instagram.

Slow entry to product pages and cart

The product card and cart are places where we usually collect the most tools: installment calculators, review modules, timers, up-sell recommendations. This makes them simultaneously the heaviest and slowest subpages in the entire service. If you see an unnaturally large “drop-off” in the purchasing funnel when moving to these pages, the problem is probably technical glut.

Good campaigns, weak sales

This is probably the most painful symptom. Reports from Google Ads or Meta Ads show wonderful results, the agency boasts of cheap clicks, and at the end of the day, you have a fraction of the sales you expected in the store’s panel.

Insight: Customers often look for blame among ad specialists, consider changing prices, or rebuilding the entire offer. But the truth is much more prosaic. Sometimes the problem is not the ad, the offer, or the price. The problem is that the store simply loads too slowly. Even the most warmed-up and ready-to-buy customer will not leave money with you if the infrastructure makes it impossible for them.

Is every widget bad? No, but not every one is designed for e-commerce

Reading about all these threats to Core Web Vitals, you might get the impression that I am an absolute opponent of any additions on the site. Nothing could be further from the truth. Effective e-commerce widgets are powerful revenue generators. The problem is that you will find thousands of plugins and scripts on the market that were designed with simple blogs or business cards in mind, and not advanced sales environments.

Difference between function and implementation

Just having a function (e.g., a bar with free delivery) is not enough. What matters is how this function was coded. You can have two seemingly identical popups from two different providers. One of them will load asynchronously, “weighing” only a few kilobytes and in no way burdening the browser. The second will download an entire powerful jQuery library, block site rendering for 3 seconds, and destroy your LCP indicator. The function is the same. A chasm divides the implementation.

A good widget should earn, not just “look”

In e-commerce, we too often succumb to the temptation of form over content. We install tools with plenty of unnecessary animations, flying confetti, or heavy fonts that must additionally load. Remember: the customer does not enter your store to admire the window closing animation. They enter to buy a product. A good widget is meant to help them in this decision, not distract them and delay the process.

Technical lightness becomes a competitive advantage

In the SEO and CRO environment, we coined the saying that “speed is a new functionality.” Your fast online store has a huge advantage over the competition, which bends under the weight of a dozen plugins. If you use light, well-optimized widgets, your site ranks higher, and customers go through the purchasing funnel smoothly and without frustration.

Why “more features” does not always mean “better”

Are you tempted by “all-in-one” tools that offer hundreds of ready-made mechanisms, from surveys to complicated gamifications? From a technical point of view, this is a trap. Even if you use only 5% of the capabilities of such a gigantic tool, your site often must load the code responsible for the entire 100%. It is exactly this redundant, unused JavaScript code (so-called unused JS) that chokes performance the most.

Why DropUI has the advantage when Core Web Vitals and sales matter?

Seeing the scale of this problem during my analyses, I knew that e-commerce owners need a change in approach. They want to have effective marketing tools but cannot afford slow page loading. It’s not a matter of choosing between marketing and technology. It is precisely at the junction of these two worlds that DropUI was created.

As an expert, I will not just tell you: “DropUI has cool widgets.” There are hundreds of tools on the market that have widgets. I will say what, from the point of view of your business and performance, has real meaning: DropUI solves the problem of redundant, heavy, and scattered implementations.

How exactly do we protect your fast online store from a drop in results?

One system instead of several heavy applications

Imagine that instead of a separate tool for popups (script no. 1), an application for sticky bars (script no. 2), and a system for exit-intent (script no. 3), you have only one, optimized code. By consolidating these mechanisms under the DropUI umbrella, you drastically reduce the number of requests to external servers. Your browser downloads data once, not three times.

Fewer scripts = more control over performance

By uploading subsequent widgets to an online store from different providers, you lose control over what is actually happening on your site. In one centralized system, you know exactly how the tool affects the site architecture. This makes LCP optimization and the control of other indicators incomparably easier and more predictable.

Widgets designed for conversion and speed, not just appearance

At DropUI, we do not add unnecessary, heavy bells and whistles to the code. We focus on what actually brings profit. Our solutions are coded asynchronously using modern standards. This means that the widget waits patiently in the background and fires exactly in that fraction of a second in which it is needed, not blocking the loading of the main content of your store.

Easier maintenance of technical order in the store

How many times during audits have I found “remnants” of scripts in the code of stores after agencies with which the company has not cooperated for two years? Scattered implementations are a guarantee of technical mess. By using one, efficient environment, you take care of digital hygiene. When you manage everything from one place, turning a given campaign on or off does not require developer intervention.

Lower risk of “hidden performance debt”

Performance debt is created slowly. You plug in a small code today, another in a month, and after a year you realize that the store works half as fast. By choosing the DropUI architecture, you minimize this risk. Instead of multiplying problems, you choose a solution in whose DNA supporting conversion without causing damage to your SEO has been written from the very beginning.

Core Web Vitals vs CRO - does a fast store really sell better?

When I talk to e-commerce directors, I often see an artificial division. On one side of the barricade stands the SEO department fighting for milliseconds, and on the other - CRO (Conversion Rate Optimization) specialists who want to add subsequent cart-saving popups. Meanwhile, the truth is that technical page load speed is the foundation of effective CRO. Let’s see how these two worlds connect and why a fast online store is a machine that simply sells better.

Speed affects the first impression

In sales psychology, there is a phenomenon called the halo effect. The first impression influences the evaluation of the entire assortment. According to industry reports investigating consumer behavior (e.g., analyses published by Akamai), a user forms an opinion about a store in just 50 milliseconds! If the page loads slowly and the images “jump,” the customer’s subconscious sends a signal: “This store is amateurish and does not inspire trust.” If you want to sell premium products, your site must work in a premium standard.

The mobile user is less patient than you think

The majority of traffic in e-commerce today comes from smartphones. And a mobile user is a user on the run, often on a weaker connection (e.g., in public transport). Mobile behavior studies show the brutal truth: if a site loads longer than 3 seconds, more than half of the visitors simply leave it. Every additional second of waiting is a drastic drop in the chances that the store’s conversion will end in success. On mobile devices, patience does not exist.

The widget should help buy, not prevent from buying

Marketing tools are great sales assistants, but only when they know their place. Imagine you enter a physical boutique, and a seller literally throws themselves at your neck shouting about a promotion, while simultaneously covering the clothes rack. That is exactly how a sluggish popup works, blocking the LCP load. A good widget should gently suggest and point out benefits, not become a technical obstacle on the way to the checkout.

The best UX is the one the customer does not notice

From the point of view of experience design (UX), the best-optimized interface is “transparent.” The customer does not think about how quickly the page loaded or how smoothly the menu expanded. They simply browse the assortment, add to the cart, and pay. When you eliminate lags, delays, and jumping elements, you remove cognitive friction. It is then that your marketing activities fall on the most fertile ground.

How to make a fast online store without giving up on marketing?

I hear this question most often when my SEO audit lands on the table. “Do I have to remove all popups and recommendations now to satisfy Google?”. My answer is always: absolutely not! Online store optimization is not about cutting marketing to the root. It is about intelligently building the technological environment.

It’s not about “less marketing,” but about smarter implementations

You don’t have to give up on saving abandoned carts or informing about free delivery. Instead of cutting functionalities, you must change the way they are served. A smart implementation is one in which the code of marketing tools loads asynchronously, conditionally (only where it is needed), and without pulling unnecessary libraries from outside. It’s about code quality, not about giving up on sales funnels.

Better tool architecture = better performance

Instead of installing five different plugins from five different developers, each of which pulls its own heavy tracking scripts, bet on integration. I have already mentioned this in the context of DropUI - one well-designed system that handles popups, bars, and recommendations will always win performance-wise with a mosaic of accidental applications. Better architecture means cleaner code, less server load, and a peaceful head of the owner.

The best stores do not choose between SEO and CRO - they combine both

E-commerce leaders know that traffic without conversion is burning money, and conversion tools without traffic (because the store dropped in Google due to poor Core Web Vitals) are useless. The best brands treat technical SEO and CRO as one, cohesive ecosystem. A fast, optimized store attracts cheap and high-calorie traffic from search engines, and light, intelligent widgets turn this traffic into buyers.

Performance is a part of the sales strategy today, not a task for a developer

We must change the way we think. Marketing directors and business owners often shift the speed problem onto the shoulders of programmers, expecting them to magically “accelerate the server.” Meanwhile, it’s the marketers who are responsible for plugging in subsequent scripts! A fast online store is a common goal of the entire team. When you understand that performance directly drives your ROI from the campaign, technical optimization will stop being an unpleasant duty and will become your most important sales strategy.

Checklist: Is your store overloaded with scripts?

Are you wondering if the performance problem applies to your e-commerce? Let’s do a quick examination of conscience. Go through the checklist below. Answer honestly to 5 questions:

• [ ] 1. Do you use more than three separate external tools for on-site marketing (e.g., a separate provider for popups, a separate one for chat, a separate one for recommendations)?

• [ ] 2. Does your store’s score in the Google PageSpeed Insights test on mobile devices for the LCP (Largest Contentful Paint) indicator light up in red or yellow (above 2.5 seconds)?

• [ ] 3. When entering your store from a smartphone, do you have to close two windows (e.g., cookie bar + immediately popping up signup popup) before you see the product?

• [ ] 4. Do you yourself feel a so-called “freezing” of the screen for a fraction of a second when you try to expand the menu or click on a different product size on the phone?

• [ ] 5. Are you sure that there are no “remnants” of scripts in your store’s code after advertising agencies or applications that you gave up months ago?

Result: If you answered “YES” to point 1, 2, 3, or 4, or “NO” to point 5 – your store suffers from performance debt. You are losing customers you acquired with your own advertising budgets.

Mini Audit: 5 signals that your store is slower than you think

Often it seems to us that our store works lightning fast, because… we visit it every day, and our browser has saved files in the cache. However, customers see it differently. Here are 5 hard business signals that I pick up during audits:

1. Growing Bounce Rate from paid campaigns: CTR from ads on Facebook is high, clicks are cheap, but the bounce rate on the store exceeds 70%. Users click but close the page, unable to wait for the loading of the assortment blocked by scripts.

2. Chasm between Mobile and Desktop conversion: If 3% of users convert from computers, and only 0.5% from smartphones, it’s not the fault of the fact “that people don’t buy on phones.” Phones have weaker processors that simply cannot handle the grinding of heavy widgets.

3. Customers complain about “escaping” buttons: If the customer service department hears that someone wanted to click “Add to cart” but hit the regulations, you have a critical problem with the CLS indicator (most likely due to bars and popups loading with a delay).

4. Analytics “loses” traffic: You see 1000 clicks in the Meta Ads panel, but Google Analytics registers only 700 entries. Why? Because the Analytics script loaded too late (after heavy popups), and the user managed to leave the page before the system counted them.

5. Good SEO, weak traffic: Your assortment and product descriptions are great, but for some reason, you can’t break into the TOP3 in Google for key phrases. Since the Mobile-First Indexing and Core Web Vitals updates, Google artificially stifles the visibility of stores that have poor performance.

In e-commerce, the winner is not the one with more widgets, but the one with better performance

From my perspective, i.e., a person who connects the world of technical SEO with marketing automation on a daily basis, the conclusion is extremely simple. In e-commerce, the arms race for the number of pop-up windows, flashing bars, and spinning wheels of fortune ended long ago. Today, the winner is the one who can provide the customer with a smooth, trouble-free, and lightning-fast shopping experience, in which marketing wisely supports the decision, rather than blocking it.

If you were to remember only a few of the most important things from this article, let them be these final conclusions:

• Widgets can greatly increase conversion, but only when they do not slow down the store. A tool that makes you wait for it closes the customer’s wallet before they manage to get acquainted with the promotion.

• Core Web Vitals is not a “technical detail” today, but the foundation of the sales strategy. Page load time and interface responsiveness are the same business indicators as ROAS or the size of an abandoned cart.

• LCP, INP, and CLS have a real impact on UX and SEO. Sluggish scripts destroy these parameters, which directly translates into drops in Google and the frustration of buyers.

• The best solutions are those that are light, well-thought-out, and integrated. Choose systems (such as DropUI) that handle many marketing functions from the level of one optimized code and do not multiply unnecessary scripts burdening your customers’ browsers.

When you next consider implementing another free plugin for collecting emails or a timer, remember this one principle:

The most expensive widget is not the one you pay a subscription for every month. The most expensive widget is the one that, poorly optimized, quietly takes away your most precious traffic and eats your sales.