Website Performance is Marketing Performance
It sometimes seems like everyone's obsessed with performance these days. And why not? In the data age, we have the tools to analyze and improve processes in every area of our lives, from health and fitness to productivity and relationships.
We're obsessed, too. Not with fitness optimization (though maybe we should be) or even with time management optimization (though we definitely should be). We are obsessed with website optimization. Hopefully, by the end of this article, you will be too.
What is website optimization? Why should you care?
Years ago, I read something that stuck with me: there's no such thing as "more optimal." Optimal is binary; you're optimal, or you're suboptimal. Optimization is a march toward an ideal form. It's a series of attempts to reach perfection, and let's be honest: you won't.
Website optimization is ultimately about improving outcomes. This article will argue that best practices for technical performance naturally result in improvements to other areas. By optimizing your technical performance, you will see improvements in your marketing performance indicators. We will prove it.
Technical website optimization can get complicated, surprising nobody, but some of its principles are easy to understand.
Just as there are many ways to skin a cat, there are many ways a web developer can code the same web page. Visually, two web pages can look identical. But not all methods of web development are created equal. Some websites are inefficient. Like a high school essayist trying to reach the minimum word count for an assignment, some websites use more HTML code than they need to.
Over the course of a whole web page, this kind of inefficiency has a tangible effect on website speed: it increases file sizes. It also simply takes more processing power to render the same design. Your device has to read every part of the code, so the more parts you include, the slower the page becomes.
Most website designs rely on media — photos, videos, illustrations, etc. — to grab attention and make the page attractive and memorable.
Just like code, media can be efficient or inefficient. Through the black magic of media encoding (don't ask us how it works — really, don't ask), we can reduce the file size of an image by 90% while barely impacting its quality.
When we upload full-sized images, videos, and other media without optimizing it, we can wind up adding a massive amount of data to each page. A 500kb web page can very quickly become a 15Mb web page. That's no bueno.
When you visit a website, you connect to a server. Because servers are computers, they can be faster or slower depending on their hardware. Suppose your server has about as much RAM as an original 1985 Nintendo Entertainment System. In that case, you will not have a fast website. No amount of code optimization or image encoding will make your website fast because the server acts as a bottleneck.
For many years, Google has maintained a tool for web developers to test their website performance. It's one of many available, but it's the one we'll focus on. Why? Google uses PageSpeed, particularly mobile PageSpeed, to determine their search rankings and analyze their advertiser's websites. Excellent PageSpeed scores can drop your advertising costs and boost your SEO performance, as we'll soon see. Terrible PageSpeed scores can, as you may imagine, do the exact opposite. If your website has a lousy PageSpeed score, you're bleeding money on your ad campaigns and losing organic search rankings.
How does it work?
The PageSpeed analysis tool loads your web page on a simulated mobile phone and a simulated desktop computer. As the web page loads, the tool checks for a variety of performance indicators. The mobile device simulation is handy, as it simulates a device with a poor connection and less memory. It means that problems become more apparent on mobile devices — a scenario that plays out in real life all the time.
How PageSpeed Impacts Your Bottom Line
Let's explore some business cases for optimizing your website performance.
The Case for SEO Performance
Google uses PageSpeed as a key indicator for search rankings. Fast sites get favourable positioning versus equally relevant, slow websites.
Is it the most significant factor in SEO? No. Is it an important factor? Yes, definitely. If all else stayed the same but your website loaded one second faster, that performance improvement would bear out in your rankings over time over time.
The Case for Advertising Efficiency
Google Ads is one of the biggest and most commonly-used advertising platforms in the world. When you advertise on search engines like Google, your pricing is adjusted based on the quality and relevance of the advertisement and your landing page (or website). Google calls this a 'Quality Score.' Low-quality scores artificially reduce the efficiency of our ad budget — you pay the penalty for poor relevance and poor quality. On the other hand, exceptionally high-quality scores result in a pricing discount, stretching your budget further.
When we've worked to improve website performance for our advertising clients, we have consistently seen an increase in quality score, even if our keywords, design, and advertisements stayed the same. This alone provides a tangible increase in advertising performance.
The Case for Marketing Performance
Improving page speed and user experience has a knock-on effect for all digital marketing campaigns.
Marketers and business owners alike tend to over-value their offering. That is, we tend to expect that people want what we're selling with more urgency than they sometimes have. Especially in the digital arena, clicks are cheap, and attention spans are short.
That makes website performance surprisingly impactful on marketing conversion rates. Take a look at CloudFlare's report on website performance versus conversion rates:
In a nutshell: speed = conversions.
The Case for Sustainability
Finally, website performance impacts the carbon footprint of your business.
Web pages have a measurable CO2 footprint. No, really. Go measure your website. Reducing the footprint of your web pages has a multiplicative effect on emissions. We recently launched a microsite for a client that produces just 0.34g of CO2 per pageview versus just over 1g of CO2 on their main website. In just a few weeks, we've reduced their advertising emissions throughput by approximately 50%. We estimate that our microsite alone saves 4.3kg of carbon per year compared to running ads through the main website. It's a small but measurable impact. If every business prioritized web performance, particularly website sustainability, we'd see significant global reductions in IT-related emissions.
PageSpeed Impact on Advertising: Real-World Data
For business confidentiality reasons, we've omitted the client name and identifying details from this article. They are a B2B services provider whose website primarily focuses on lead generation.
Our client's services became particularly appealing for many businesses in March 2020, as COVID's first wave profoundly modified our way of life. They were able to pivot to serve the explosive demand for PPE products. During this time, lead generation was incredibly affordable through paid search advertising.
As society adjusted to our 'new normal' (we shudder to use the term), two things happened that negatively affected our advertising efforts. First demand cratered, as previously overwhelmed supply chains caught up with COVID realities. Second, competitors who were not previously advertising had every incentive to catch up. The search advertising landscape became more competitive across every vertical as more businesses moved online. By mid-2021, lead costs had skyrocketed while lead volume worsened, a losing combination.
Over time, our advertising department realized that website speed might be impacting performance.
We noticed that since integrating HubSpot's on our landing pages, PageSpeed scores had dropped from an acceptable range to 24/100 on mobile devices. Our hypothesis was that slow load times impacted both quality scores, flagging our most popular keywords and overall conversion rates. We could test this hypothesis by improving the PageSpeed score.
We knew that the website had previously performed well in PageSpeed tests. The only speed-related variable to change over the past several months was the HubSpot integration. HubSpot's tracking script loads several additional scripts that significantly slow down web pages. It is not possible to reduce the impact of these scripts directly. Instead, we chose to bypass HubSpot by introducing Integromat.
Integromat is a Zapier-like tool that glues web services together with workflows. For our scenario, we integrated Netlify forms with Hubspot using Integromat. Netlify's free form handling service could now post lead data directly to Hubspot without a bulky third-party script.
We improved our score from 24/100 on mobile to 72/100 - a definite improvement, though still less than ideal.
We noticed an immediate uptick in some performance metrics without significantly altering campaigns, keywords, ads, or targeting settings. Our click-through rate improved by 14%, and our search placements improved. In the thirty days before making PageSpeed improvements, we had a top-of-page rate (at which our ads appear anywhere in the top set of results) of 79%. After optimization, our top of page rate improved to 84%. Our absolute top of page rate, which measures how often we have the very top position, improved from 62% to 72% over the same period.
Appearing as the very first result significantly more often had a massive effect on conversion rate. The previous thirty days had generated 17 quote form submissions. After optimization, we generated 23 — a 35% improvement in lead generation.
We wanted to push performance even further, but we knew that the HubSpot integration meant that we could not improve it further.
Instead, we created a marketing micro-site. Our team identified an open-source HTML template that looked similar to the main website, then edited it to conform with brand guidelines more closely. Each page provided product information and CTAs. Each was linking to the main website's quote form. We kept integrations to a minimum to focus on speed. Using this technique, we improved mobile PageSpeed scores on our landing pages from 72/100 to over 90/100 on average.
Upon further improving average PageSpeed, we saw a more significant increase in advertising performance. We can guarantee the previous performance boost was due to improved website performance as little more than regular campaign maintenance occurred during that period. But the microsite introduced significant changes to the marketing funnel that may muddy the cause of our results. Nonetheless, we can confidently say that improved PageSpeed was one of the essential factors in improving performance.
Upon launching the microsite, our cost efficiency improved. Our click-through rate continued to improve, increasing by 2%. Our cost per click declined by 23%. Most importantly, we further enhanced lead generation. Through the past 20 days since implementing the microsite, we've seen thirty-five leads from advertising. Comparing this to the previous period (after initial optimization but before the microsite), we generated 26 leads. This amounts to a 35% increase in lead generation at the same cost per lead, allowing us to scale the campaign effectively.
The bottom line
Web performance is marketing performance. Marketing performance is revenue. In some cases, improving website performance will passively enhance your lead generation through every channel due to higher visibility, lower advertising costs, and better conversion rates.
If you're looking for easy wins for your digital marketing, consider performance optimization. It's the rising tide to lift all ships.