Want people to stay on your website longer? 5 things to check asap

On Wednesday evening, right when I was about to wrap it up for the day, I got an email from my client saying that the new copy I wrote for them increase the bounce rate* for their homepage from 2.3% to 81.1%. 😱 * - bounce rate: a % that tells you how many visitors visited only that one page and left without visiting a second page (i.e. bounced off it) Which meant that before my changes, only 2 out of 100 visitors were leaving their site without checking another page. After my changes, 81 out of 100 left without exploring more. 81 freaking visitors out of 100! On a scale from 0 to 10, how much do you think I panicked? Zero. Ok, maybe one. And you know why? Because they weren't seeing any reductions in conversions, and the leads were better qualified (I asked). That's when I knew that my first suspicion was right: That initial 2.3% bounce rate? A measurement error! But how do you prove that it's an error? I can't possibly say, "There is no way in hell that out of 100 people who come to your website 98 decide to stay." and hope they'll be happy with that explanation. (Although I guess we can all agree that is as likely as you winning a lottery tomorrow). So I put my investigator hat on and dug into their website performance data. Bounce rate too high? Here's what to do 1. Make sure there is no measurement error It's hard to say what a good bounce rate is because it depends a lot of the page, your business model, and your traffic sources. But I'd say a bounce rate below 20% for a homepage is super suspicious. And 2.3%? No way in hell that's accurate. How to know for sure? Compare the bounce rate you see under Behavior -> Landing pages with the drop-off rate you see under Behavior -> Behavior flow. They should be very similar (although these are technically different metrics). In case of my client, the drop-off rate for the old copy was 81.1% vs the bounce rate of 2.3%. So yeah, definitelly an error. 2. Check your page loading speed Go to [Google PageSpeed Insights](=.LZi_BEbyCIcjTYFpCH-hPtWaSa64oeqcknngylUO4B8), copy-paste your URL, and wait a sec for their neat report. Ideally, everything should show up "green". If not, the report will tell you what issues there are and how to fix them. If you see that the load speed on mobile is slower than on desktop (or vice versa), look at the bounce rate report for different devices: - Go to Behavior -> Landing pages - Click on the URL of a page you're analyzing - Select "Device category" as secondary dimension In case of my client, the PageSpeed report showed that the page loaded too slow on mobile. Their mobile visitors currently have to wait 12 seconds (!) till the page becomes interactive. The Analytics report showed that the difference between the bounce rate on mobile vs desktop was 89% vs 64%. Although some difference in bounce rates is typical (the bounce rate on mobile usually higher), 27% of a difference is too much. Combined with the data from the PageSpeed report, we could now confidently say that the load speed was impacting the bounce rate. 3. Look at your heatmaps & session replays Are you recording heatmaps and session replays? I hope you do (if not, [Hotjar](=.JB0Ss6dYkTH51e-9-drYZV9mgppP7ILO93xysmRrse0) has a nice free plan and is super user-friendly). Look at the session replays on different desktop devices (for starters) to answer the following questions: - Does all the banner content fits within above-the-fold space on a typical laptop (not the giant Mac) screen? - Is the banner designed in a way that the visual hierarchy guides the visitor from important parts to less important? - Does the text look easy to read and quick to grasp? Or does it look like reading it means work? Most website visitors bounce off your homepage because the banner failed to communicate relevancy and value fast. In case of my client: - The homepage banner was cluttered (too much text on too little space) - The content that was supposed to be visible above the fold didn't fit there because of how the space was used - The page had too much unnecessary white space that made it longer than it should be I didn't consult them on design before the pages went live but now I probably will :) 4. Create a prioritized list of changes Especially if you discovered too many problems, don't try to tackle everything at once (hello, overwhelm!). Start with the low-hanging fruit: something that's easy to implement but promises a big reward. In the case of my client, I recommended doing these things first: - fixing the load speed on mobile - changing the banner And also checking Google Analytics installation to make sure the data can be trusted. Which is not really a low hanging fruit but something that must be done before anything else, or we'll have no reliable data. 5. Make the changes. Measure. Analyse. Repeat. Once you made the changes, don't forget to reset your heatmap so it shows only the data after you made the changes. Repeat the steps 1-4. I can't say what will happen to my client's homepage. I sent them an email with my analysis literally an hour ago. But I'll keep you posted. (And yes, I write my weekly emails on a VERY short notice 😄) This would be all from me for the week. Rainy greetings from Germany, Gill 🌐 [Get feedback on your website](=.H2nQj6cGV1XEQBev0xn5NO6mqmJbAZDIfVNoFPkzAJw) | ✍️ [Get new copy](=.LJYQTuimK_fmVbyLgksBRP9K6UqMdGhQcxFK8pOcjuM) | 🖥️ [Find your UVP](=.1__E3PUTpUczcYor-XogurQJchHu3toCtWFlU7ANjPU) P.S. This email may contain typos, and I'm fine with them because cloning humans is, unfortunately, still impossible. Spending more time proofreading my emails would mean I'll have to share fewer tips with you. And sharing more and better tips is more important to me than sharing tips that are grammatically perfect. I hope that's fine with you, too. [Gill Andrews] [gillandrews.com](=.4lYVja9RtNvoqcyhHoG1_ElLuTU_tJ2fpzmimKfLXI0) [Share to Facebook](=.gVk6A300DvDtO3IaVzm3b2FPrOppfhf7t9zyecbWwaY) [Share to Facebook](=.sLqFV2L1z-lWX7P1bjnI90v9fh5f2NExBeImJm0_grE) [Share to Twitter](=.GOQbrV4COVAnhUIxD8CZAxO0tqJMyUrvR0Y5hb7fRRA) [Share to Twitter](=.mhZjQhgwi5DL9u1U1VctFcByRNPUSloUIqA8gL229BY) [Forward email](=.8jesvLQrA3640JpC34O7Rv4KnwkRSFVdQPLOJTF1_Ek) [Forward email](=.L828XPf1VHHdzZYdpSWSBw5fKWqjKBOxJNbylJVof9A) Gill Andrews You received this email because you signed up to my updates. Changed your mind? No problem. Unsubscribe using the link below. [Unsubscribe](=.gaEDiAEXa14BNuWvBIaxMysbHV9QKVfp5i7VZaIDbQE)