Updated: March 2020
There are no affiliate links on PagePipe.
Remove third-party services like Hubspot, Jivo chat and Optinmonster. They are much more significant than font removal for speed improvement. They’re heavy and slow – and thus low-hanging fruit. Pareto’s law (80/20) of focusing on these assets gives a bigger return on investment.
Chat adds 1 to 10 seconds of page load time site wide.
So it gobbles up most of the performance budget.
Questions to ask yourself about chat:
- Why do we need these plugins in the first place?
- Do we know they get real results – or we hope they will?
- If they get results, how significant is their contribution to profits?
Sometimes design compromise is a value call only you can make.
We’ve written about OptinMonster. We challenge the value of a feature with such an obvious poor user experience. It’s a distraction and intrusive. We’d like some serious fact-checking. Where are the scientific data purporting big improvements in signups from popups? Are these facts from monetized plugin affiliates? We bet audiences change their web behavior and ignore or dismiss popups. Some swear under their breath when unwanted screens pop up. They interrupt a search for problem-solving content.
Can you find ways to combine off-site services? For example, if you use Hubspot, they also have a chat function. Send one request instead of two for services. Value analysis is an industrial discipline using combination, simplification, elimination, standardization, and substitution.
When you do value analysis to make speed decisions, you’re being creative. Can you do “combination” of functions to lighten the load? That’s good problem-solving. If loading something heavy like Hubspot, see if you can get more mileage from it. Will it be lighter to use their livechat instead of adding another service like Tawk.to or Jivochat?
Do we have a recommendation for a lighter-loading live chat function?
Live chat is a bane for speed. We haven’t researched it much because it’s so awful. We think it’s going to die. Holler Box plugin has a “faux chat” function. We’ve found it a clever and fast alternative. But it’s not “live chat.”
Chat is very trendy and faddish right now. Much like sliders were in 2012. Both chat and sliders are bad for UX. Documented with user surveys and metrics, sliders prove inefficient. But chat hasn’t yet. It will be unpopular as people become blind from overexposure and poor results.
We have chat-insistent clients who get no benefit in profitability. Their justification? They want chat to stay on their site because “everybody is doing it.” What did your mother teach you about peer pressure?
So look hard at the benefits (profit value) of these heavy non-features. Are there real numbers proving they contribute? Or just wishes and whims?
Providing chat service creates a customer expectation of no lag or wait time. Provide answers fast. If they’re lame or slow, then you create a negative user experience instead of a positive one. A negative is a liability.
|Name||FCP||JS Time||Weight||Chat displayed|
If MailChimp messes up and doesn’t deliver your free report, email us and we’ll kick their monkey butt.