Most SaaS founders think they have a traffic problem. I kept seeing something else.
People were clearly interested. They spent minutes on the pricing page, hovered the CTA, scrolled back and forth, and then just left. No signup, no chat, no feedback.
I tried chat widgets. They didn’t help. Hardly anyone clicks them, and the people closest to converting almost never do. By the time they hesitate, they’re already leaving.
That made me question the model. Why are we always waiting for the visitor to start the conversation?
So I built Concier.
It watches how visitors behave in real time and steps in when the intent is high. Not with a generic “how can I help,” but something tied to what just happened.
“Figuring out which plan fits how you'd use this?”
That one question surfaces objections you’d otherwise never see. Pricing concerns, missing integrations, hesitation you can actually act on.
It’s just a small script you add to your site.
Now I’m looking for honest feedback from people with real traffic.
Would this feel helpful at the right moment, or just annoying? Check it out here - https://tryconcier.com
This is interesting.
Went through the page. The problem is clear and hits hard, but the first impression leans more on the pain than the outcome.
It shows what’s happening, but not instantly how this is different from the chat widgets you mentioned.
Might be why some visitors hesitate even if the idea is strong.
You're right. The page explains the problem well but doesn't immediately show why this is different from just another chat widget.
That's the thing it needs to land in the first five seconds. Going to fix that. Thanks for actually looking.
I like this distinction. Traffic is easy to measure, but it can hide a weaker problem: the visitor doesn’t immediately recognize their own pain in the product.
For technical products especially, I think the useful test is not “did they click?” but “can they name the exact workflow they would use this for?” If they can’t, more traffic just creates more vague interest.
Have you found a good way to separate messaging problems from channel problems when you review SaaS sites?
The workflow test is a good one. If someone can't describe the specific moment they'd use it, the page is speaking to a category of problem not an instance of it. And people buy instances not categories.
The way I've started thinking about separating messaging from channel is looking at where drop-off happens. If people aren't clicking through at all it's a channel problem, wrong audience or wrong hook. If they're landing and spending time but not converting it's almost always messaging, the page isn't connecting what they already feel to what the product actually does.
The tricky middle case is when time on page is high but intent is low. Looks like engagement, reads like interest, but they're actually just confused. Session recordings help there more than any metric does.
Feels like the issue isn't just traffic or timing , it's clarity at that moment.
If users hesitate on pricing and don't instantly see why one option fits them, they just leave instead of asking.
That's exactly it. The hesitation isn't confusion about the price, it's uncertainty about fit. Which plan is actually right for my situation. And most pricing pages answer that with a feature comparison table which doesn't help because they still have to map the features to their own context themselves.
The question nobody asks out loud is usually "is this the right one for me" not "what does this include."
I would try it
I’ve enabled a free month on your account, just install the script tag on your website.
oh thanks, ill give you a honest feedback once i try it
Sure thanks
Awesome, really appreciate it. Just sign up at tryconcier.com and reply here with your email or DM me. I'll make the first month free and help you get it set up personally.
The "wait for them to click" model breaks at the exact moment you actually need it — your data on the hover-and-leave pattern matches what we see at TokRepo (~38% of visitors who hover the pricing CTA for 6+ seconds bounce within 30s if no proactive engagement happens, vs ~12% if they get a contextual prompt within the first 4s of hesitation).
Two notes from running this in production for 4 months on tokrepo.com:
The trigger window is narrower than people expect. Anything past ~5 seconds of pricing-page hover and the proactive prompt feels like surveillance, not service. We A/B'd 3s/5s/8s — 5s won by 18% on engagement-no-annoyance composite, 8s tanked NPS by 0.4.
The phrasing matters more than the timing. "Figuring out which plan fits how you'd use this?" works because it names the actual cognitive task. Generic "any questions?" tested at 2.1% reply rate vs 11.4% for task-specific. ~5x.
The "intent" signal you really want isn't dwell time alone — it's dwell-time + scroll-pattern (back-and-forth between tiers = comparing, single direction = scanning). The compare pattern converts at 4-7x the scan one. That's the slice worth interrupting.
Would test on tokrepo if you have a dev sandbox — happy to share the segment data we already have on hover patterns vs paid signup.
This is genuinely the most useful comment I've gotten. The 5s sweet spot matching what we landed on after testing makes sense, and the NPS drop at 8s confirms what I suspected but didn't have data for yet.
The phrasing finding is exactly what someone pushed back on earlier in this thread and your numbers make the case better than I did. Task-specific framing at 5x generic is hard to argue with.
The scroll pattern signal is the one I haven't built yet and you're right that it's probably the most valuable slice. Back and forth between tiers is a completely different intent signal than a single pass. That's going into the next version.
Would genuinely love to see the segment data.
I think that's a great idea. I'm experiencing the same issue. I don't have a lot of traffic, but, for the amount of traffic that I do have it's troublesome that nobody signs up for email or even for the free tier of my website.
That's actually the exact situation Concier is built for. Low traffic makes every visitor more valuable, not less, because you can't afford to lose the ones who were genuinely interested.
If you want to try it, sign up at tryconcier.com and drop your email here or DM me. First month free and I'll set it up with you personally.
I’ve seen this too — people don’t need help earlier, they need it right at that “almost click” moment.
If the message feels relevant (not generic), it actually pulls real feedback instead of just annoying users.
Concier is the right kind of name for this.
The trust leak is the “try” wrapper.
That usually signals temporary product, early experiment, or placeholder brand — which is the worst possible frame for something built around catching hesitation and converting intent.
The product is trying to reduce buyer uncertainty.
The domain should not introduce its own.
That's a fair hit and honestly hard to argue with. The name communicates exactly what the product does. The domain undercuts it.
Noted. Genuinely.
Exactly.
“Concier” is doing the right job.
The issue is that “try” makes it feel like the product is still asking for permission to be taken seriously.
For most tools, that’s fine.
For something built around reducing hesitation and converting intent, it’s a bigger problem because the brand has to model the trust the product is promising.
If the product is about removing uncertainty, the domain should feel certain too.