We removed the Intercom chat bubble and had MORE conversations— not less

Last month I was getting annoyed by our Intercom live chat bubble. I decided to remove it and replace it with a link to live chat in our navigation.

It lead to a 62% increase in new chat conversations:


I find those live chat bubbles annoying— so this felt like a win.

  1. 16

    I like to see this!

    Hopefully it will inspire others! I cannot stand live chat bubbles anymore. 😬

    1. 4

      I never understood this trend... like throwing a pop up in your face for a newsletter. Hate it.

      1. 4

        A real "potential" customer is usually interested in talking and businesses only care about customers. Everyone else is irrelevant and can be annoyed endlessly haha.

        1. 4

          This comment was deleted 8 months ago.

          1. 1

            Looking up Plausible Analytics based on this post. Thx

      2. 1

        Exactly. To a lot of "marketing" people you are just a number. If they annoy 30% of people, but make 3 sales from the rest, they put anything down to your throught :(

    2. 1

      I hope this turns out to be a definitive nail in the chat bubble grave.

    3. 1

      same, those chat bubbles get in the way 99.99% of the time. I sometimes click it looking for a way to close it.

    4. 1

      Thanks Rosie! BTW: Any reason why this post didn't make the front page? It has 6 votes but I can't seem to find it anywhere on the first, second or third page 🤔

      1. 1

        Not all posts make it to the homepage. Often ones with upvotes do, but not necessarily straight away.

        There have been some changes, Courtland wrote this - https://www.indiehackers.com/product/indie-hackers/changed-homepage-algorithm--MDwO4w1vQTbkQmSzvmH

  2. 2

    Hmmm, this is interesting. Personally, I hate all those bubbles.

    But as for the results... 0.68% vs 0.42% is just 0.26% but no 62%. And see, you have many more visitors via bubble. It could mean, for example, you have a bot there and people don't like to deal with it. So, I would leave both and investigate more carefully what's wrong with the bubble as for me - it could lead to more conversations but for some reason, it didn't.

  3. 2

    really well edited video!

  4. 2

    The main problem of those bubbles to me is that they break the UI.
    Sticking to your own UI gets more engagement as you pointed out.

    Cool insight!

  5. 2

    Wow, counter intuitive at first but makes total sense. Not a fan of the bubbles myself. Thanks for sharing!

  6. 2

    Hey Steve. This is very interesting. I have often wondered whether chat bubbles were effective. To me it seemed like the new hamburger menu - for a while hamburger menus were only obvious to techies, until they became almost ubiquitous. I felt the same with chat bubbles. Perhaps they haven't become quite the standard?

    Anyway, loved the video. Everything about it - the format, the style, the music, the editing, the lot. Having looked into your other work with SiteBuilderReport I'm guessing you've got that down to a tee, but those videos must take a tonne of time to produce, and it shows in the quality. Do you have any tips on creating YouTube content? (in terms of production or in terms of effort vs benefit, etc)

    1. 2

      Thank you! I've been trying to get better at video for years now. And while I still have a lot to learn, I feel like I'm starting to get a feel for it.

      Production: Write a script and buy a teleprompter. You can buy a cheap iPhone teleprompter on Amazon for like ~$50. This has improved my videos more than any camera I've purchased!

      Effort vs benefit: For new channels starting out, it really pays to target keywords. I've seen it seed several Youtube channels now. (I should say though: My 'Chat Bubble Blindness' does NOT target a keyword. A better example of keyword targeting is my Site Builder Report channel).

  7. 2

    Looks like a hack to increase engagement! Nice

    1. 1

      That's a good way to frame it— thanks!

  8. 1

    Statistically significant, WOW

  9. 1

    I went on one startup site, I can't even remember or care what tool they were using but someone in their big brain department decided that every time a user visits a new page, we're gonna make the chat pop up and play a loud sound.


    I can confirm that clear buttons in navigation work a treat, we use a javascript pop up button too.

  10. 1

    I think live chat bubbles are great when you want persistent support such as when users are already logged in.

    For marketing or non-member questions, I think it's better to require clearer intent.

    I never find the auto-triggers useful. (Unless it's an auto-trigger welcoming someone after they signed up, in which case that's a friendly way to show them you're their to help)

  11. 1

    makes sense - those chat bubbles are so annoying.

  12. 1

    Interesting. Kind of a clickbait title though, because you still have Intercom on the site -- you just removed the default launcher.

    In my opinion, the results of a test like this would depend on the audience of your website. For a less technical audience I can see a strategy like this working out -- it's way more obvious what a link in the header called "Live Chat & Support" is about rather than an anonymous circle in the corner of the website.

    If you're targeting a technical audience, I don't think there would have been a statistically significant difference here, since most people would be familiar with Intercom and what it does.

    I'd also venture to say that you would have had less conversations if the link wasn't so long and visible -- say, if it were only in the footer or if it were buried in a menu somewhere, as support pages usually are in the actual app itself. If you have Intercom in your app, I'd be curious to see a test for your app specifically (not including your landing page), as well as where you've placed the support link there.

    1. 3

      How is this clickbait? We literally removed the chat bubble and it resulted in more chat conversations. The chat bubble is also known as the default launcher.

      1. 1

        It's misleading because it implies that you removed Intercom entirely. What you actually did was change how people started their interaction with it.

        1. 2

          It 100% does not imply that. I think you misunderstood what I meant by chat bubble.

          1. 1

            For anyone only reading your headline (and there are lots of people who just do that), it's not clear that you've simply moved the entry point for Intercom instead of removing it completely. You only mention "removing" Intercom -- nothing about "moving" it. And for many people, Intercom's "chat bubble" is synonymous with Intercom itself.

            Anyway, it's a small thing that I just wanted to point out, and I'm sorry that you took offense at that -- maybe "clickbait" was too harsh, but it's certainly still a misleading title. Your actual blog post title ("Chat Bubble Blindness") is much better.

            I noticed you ignored my other points, but those are the ones I'd be more curious to hear you address instead.

            1. 2

              “And for many people, Intercom's chat bubble is synonymous with Intercom itself.”

              What’s clear is that YOU think that. Not “many people.”

              When you call my work clickbait you are implying I am dishonestly trying to “bait” you into my post. I take offense to that characterization.

              1. 2

                You seem very defensive and cannot seem to empathize with my point of view or well-meaning intentions, so I'm not going to respond further.

                I hope you take a walk, enjoy some sun, and have a nice rest of your day, Steve.

                1. 2

                  I will defend my work if people call it clickbait :)

  13. 1

    what are your thoughts on VideoAsk? Video chat bubbles?

  14. 1

    I hate chat bubbles as "sales" on marketing materials. Especially when they pop up and make the status of your tab flash.

    Love them inside of products as "support" though once I've already made the decision to use or purchase the product.

    Too bad this wasn't an A/B test though. Can't really for sure say this was the reason things changed. You had the sample size to be able to too.

  15. 1

    That's really cool. It's why it's so important to split test this stuff, you can get really surprising results.

  16. 1

    damn I hate those pop ups too. love this!

  17. 1

    YAY! My extension (https://hellogoodbye.app) has a purpose :)

  18. 1

    nice video. love the background.

    how many users / mo does your site get?

    1. 2

      Thanks Harry! I actually had you in mind when I created the graphic in the description :)

      Below are the results of the test— is that what you're wondering about?


  19. 1

    I usually say: Stop creating iteractions without clicks, if the user doesn't click, he doestn't expect nothing, and this is the best iteraction ever, what he expects.

  20. 1

    Thanks, Steve! I will add this point to our roadmap!

  21. 1

    Interesting. Maybe users know they are bots and not really live?

  22. 1

    I have chat bubbles and I'm so happy to have this validated. I remember arguing with sales and marketing people at the previous startup I worked at telling them that it was annoying and cluttered the interface, not to mention the extra javascript downloaded. They pushed back saying that they had closed a lot of sales through the chat bubble. Nice to have a solid solution that would make both parties happy

  23. 1

    really interesting perspective! thanks for the breakdown!

  24. 1

    I'm inspired to write a browser extension to remove the chat bubbles!

  25. 1

    Yeah I just put one on my site but getting 0 conversions. Thinking of also removing it! Thanks for the motivation

  26. 1

    Thanks a lot! I will try this approach as well.

  27. 1

    Nice one. We aren't a fan of the big chat bubbles either and placed a "Chat with Us" button to open Intercom alongside all of our other help options.

    Bonus with putting it alongside other help options is that the user will often find the solution themselves without immediately resorting to opening a conversation with us – then we are just left with important conversations such as bug reports.

  28. 1

    I think it's a very good idea. Made the same changes on our website as well.

  29. 1

    I use Crisp.chat for Watermark.ink. Most of my users don't use chat, but sometimes few visitors prefer step-by-step walk through. Such things are very easy with crisp chat also Crisp gives a very nice visual world map of my visitors.

  30. 1

    Nice. I'll try that.

  31. 1

    Hi there,

    Curious as to what were your exact numbers?

    Just asking because when a conversation rate is so low, it's hard to have a statistically significant change.

      1. 1

        Interesting. It is statistically significant. Good job with the finding!

  32. 1

    On my website (presentify.compzets.com), I have both the chat bubble as well as a Support link that opens up the same chat box. Hubspot allows me to programmatically open up the chat box on link click. And, having a link definitely increased our conversations.

    1. 1

      Presentify is such a good idea!

  33. 2

    This comment was deleted 10 months ago.

Trending on Indie Hackers
Designjoy crosses $70k MRR! Someone pinch me...😱 28 comments I quit my 1+m$/year job to work full-time on my own project 21 comments How I grew my Twitter (+9K) and Newsletter (+6k) in 3-4 months. AMA. 15 comments RocketList - a collection of cloud actions for your site 10 comments How a Twitter API suspension nearly killed my startup 5 comments Show IH: 16 Year Old Builds ₿itcoinForecast To Celebrate Currency Adoption 5 comments