Why I removed chat widgets from 4 of my own sites

I run four sites — riin.eu (my B2B services), stikily.ee and radezain.ee (two small e-commerce shops), and taphi.io (a smart bracelet brand). For months, all of them had a little chat bubble in the bottom-right corner. I removed every single one and replaced them with a short video greeting from me.

I don't think chat widgets are bad. They are excellent for support-heavy products where the visitor's question is "how do I cancel my plan" or "where is my order." But on a small business site, where the actual sale is trust in a person, a chat widget asks the wrong question of the wrong visitor at the wrong moment. That is the case I want to make.

What a chat widget quietly says to a stranger

Picture someone who has just landed on your homepage. They are sizing you up. They want to know whether you seem competent, whether you seem warm, whether you are the kind of person they would feel safe paying.

Then a little box pops open: "Hi! How can we help you today?"

That sentence assumes a problem. It frames the visitor as someone who is stuck, confused, or in need of customer service. But your first-time visitor isn't stuck. They are curious. They are in evaluation mode, not support mode. The widget answers a question they haven't asked yet, and it does so with a tone that belongs to a help desk rather than a first introduction.

I noticed this most clearly on radezain.ee. RAdezain sells handmade lifestyle pieces — the buying decision is emotional, slow, and aesthetic. Visitors don't need help; they need to feel something. A "How can we help you?" prompt was the conversational equivalent of a cashier interrupting someone browsing a boutique to ask if they were looking for anything specific. Most people just leave.

What I replaced it with

On all four sites I now use a small video greeting widget — a 30 to 60-second clip of me, in the same corner where the chat used to live. On Stikily it's me explaining what is the product we sell. On riin.eu it's me saying who I work with and what to expect if you book a call. On RAdezain it's the maker behind the pieces. On taphi.io it's a quick tour of what the bracelet actually does.

This is the part where I'll mention, in passing, that the tool I use to do this is Heyly — partly because I built it to scratch this exact itch on my own sites. But the argument doesn't depend on Heyly specifically. You could record a Loom and embed it manually and get most of the same effect. The point is the format, not the vendor.

What changed wasn't dramatic on any single metric. It was a quieter shift: more people scrolled, more people clicked through to the About page, and the inquiries that did come in arrived warmer — fewer "what do you do?" emails, more "I watched your video, can we book a time?" emails. In my experience, that is what trust-building looks like in numbers. It doesn't show up as a 47% spike. It shows up as fewer tire-kickers and more pre-qualified strangers.

"But chat widgets convert" — the objection I keep hearing

The most common pushback I get is: chat widgets have data behind them. They convert. There are case studies. Why throw away a proven tool?

Here is what I think is actually happening in those case studies. Chat widgets convert well on sites that are already a strong trust fit — SaaS dashboards, support flows, post-checkout questions, B2B demo pages where the visitor has already decided to engage. In those contexts, chat removes a final-mile friction. It is the right tool.

What chat widgets don't do well is build trust where none exists yet. They are a closing tool, not an opening one. And small business websites — the marketing consultant, the bookkeeper, the boutique brand — usually have the opposite problem. The visitor isn't in the closing stage. They are deciding whether to take you seriously at all. A help-desk widget is the wrong instrument for that job.

This matters more for service businesses than for product businesses. If you run a small agency, your prospect is not asking "how does your software work" — they are asking "is this person someone I want in my corner for the next twelve months?" That question gets answered by your face, your voice, and your patience, not by a chat thread. (For service-side specifics, I've written more on trust signals for small agencies.)

The e-commerce angle is actually similar

You'd think e-commerce is the opposite case — surely an online store needs chat for "where is my order?" type questions. And yes, those questions exist. But on small e-commerce sites, particularly maker-led brands like Stikily and RAdezain, the front page visitor isn't a customer with a support ticket. They are a stranger deciding whether your brand is real.

A chat widget on a small e-commerce front page tries to handle support and acquisition with the same tool. It does both badly. Support questions can live on a /contact page or a help email. The front page should be doing acquisition work — and acquisition for small brands is almost entirely about the founder being visible. (More on the e-commerce-specific version of this in video for small online stores.)

I kept email-based support on all four sites. Customers who actually need help find the contact page, and that's fine. What I gave up was the illusion that a chat widget was helping new visitors. It wasn't. It was just sitting there, quietly framing every stranger as a problem to be solved.

What I'd tell you if you asked

If your business is support-heavy — your customers genuinely need real-time answers to use your product — keep your chat widget. It is doing its job.

But if you sell trust, if your buyer is choosing you and not a feature list, I would seriously consider a chat widget alternative that introduces you instead of triaging the visitor. A video greeting widget is one option. A short embedded Loom on the homepage is another. A live chat alternative for small business doesn't have to be another chat tool — it can be a fundamentally different format that says "here is who you'd be working with" instead of "how can we help."

That shift — from triage to introduction — is the whole argument. Everything else is implementation detail.

So here is the question I'd leave you with, the one that pushed me to rip the widgets off all four of my sites in one weekend:

What does your chat widget actually say about how you treat first-time visitors?

If the honest answer is "it treats them like they have a problem" — and you sell trust, not support — that's worth sitting with for a minute.

Frequently Asked Questions

Are chat widgets always a bad idea for small businesses?

No, and I want to be careful here. Chat widgets are an excellent tool when the visitor's primary question is a support or product-use question — SaaS onboarding, e-commerce post-purchase, complex B2B demos where the prospect has already engaged. The argument I'm making is narrower: for small businesses where the actual purchase decision is trust in a person (consultants, agencies, coaches, maker-brands), a chat widget on the homepage often does the wrong job for the wrong visitor. If most of your inbound is support-flavored, keep the chat. If most of it is "is this person legit," reconsider.

What's the difference between a chat widget and a video greeting widget?

A chat widget invites the visitor into a text-based support conversation. A video greeting widget plays a short clip — usually 30 to 90 seconds — of the founder or service provider introducing themselves, what they do, and what to expect next. The chat widget is reactive and assumes a problem. The video is proactive and builds familiarity. They serve genuinely different goals. You can run both, but on small business sites I find the video does most of the heavy lifting that small business owners think their chat widget is doing.

Doesn't removing chat widgets hurt customer support?

On my sites it didn't, but I want to be honest about why: my sites don't have huge inbound support volumes. Customers who need help find the contact page or email me directly. If your business genuinely runs on real-time support — large e-commerce volume, time-sensitive queries, an active customer base — removing chat would create a problem. Test it on a smaller surface first. You can also keep chat on internal/account pages where it's clearly serving customers, and remove it from your homepage where it's confusing strangers.

How long should a homepage video greeting be?

In my experience, 30 to 60 seconds is the sweet spot for a homepage greeting. Long enough to introduce yourself, what you do, and what the visitor should do next. Short enough that someone who clicks "play" out of curiosity actually finishes it. The goal is not to sell — the goal is to be a recognisable person by the time the visitor reaches your About page or contact form. Polish matters less than warmth. A slightly imperfect video that feels like you beats a polished one that feels staged.

Is this just an argument for buying Heyly?

Honestly, no — and I'd rather you implement the idea with a free Loom embed than buy something you don't need. The format is what matters: a short, founder-led video on the homepage, replacing the help-desk framing with an introduction framing. Heyly is the tool I built because I wanted this specifically, but the argument stands on its own. If a Loom embed gets you 80% of the way there, that's a perfectly good place to start.