.svg)
If you chart buyer conversations on our website by hour, the busiest stretch outside the working day is not the evening and it is not the weekend. It is 1am to 8am, US Eastern. Between July 21 and November 11, 2026, RepX, the AI agent on our website, logged 2,378 buyer conversations, and 51.6% of them started outside Monday to Friday, 9am to 6pm Eastern. Most of that half did not arrive after the workday. It arrived before it.
Most teams picture after-hours leads as the buyer who shows up at 9pm or on a Saturday, and they try to fix it with a later shift. I call the real problem the night shift: the hours your team is asleep, which are somebody else's working day. You can't staff your way through it. You have to answer through it.
When do B2B buyers ask questions on a website?
Here is every conversation in the dataset, bucketed by the clock on a US Eastern sales desk.
| When the conversation started (US Eastern) | Share of 2,378 conversations |
|---|---|
| Weekday, 9am to 6pm (the desk is open) | 48.4% |
| Weekday, midnight to 6am | 21.6% |
| Weekday, 6am to 9am | 15.7% |
| Weekday, 6pm to 10pm | 5.7% |
| Weekday, 10pm to midnight | 2.1% |
| Weekend, any hour | 6.4% |
Source: internal Storylane data (RepX website conversations), 2,378 conversations logged to HubSpot, July 21 to November 11, 2026.
Evenings (6pm to midnight) hold 7.8% of conversations. Weekends hold 6.4%. The weekday hours between midnight and 9am Eastern hold 37.3%, more than twice evenings and weekends combined.
The weekend buyer is real (see what weekend buyers ask), but on our site the weekend is the smallest slice of after-hours demand. The big slice is the US pre-dawn.
Does staffing both US coasts cover after-hours leads?
No. The share of conversations outside working hours depends on whose clock you use:
| Desk staffed Monday to Friday | Conversations outside desk hours |
|---|---|
| 9am to 6pm US Eastern | 51.6% |
| 9am to 6pm US Pacific | 67.8% |
| Both coasts (9am Eastern to 6pm Pacific, a 12 hour day) | 47.6% |
Source: internal Storylane data (RepX website conversations), n = 2,378, July 21 to November 11, 2026. The 95% confidence interval on each figure is about plus or minus 2 points.
A West Coast team misses two in three conversations. Adding the East Coast for a 12 hour day still misses a little under half. A second coast buys late afternoon and early evening Eastern, when buyers were already fairly quiet. The busiest hours, before dawn in New York, stay uncovered.
So the useful question is not "how late should our SDRs work?" It is "what happens to a buyer who asks at 4am?"
Who is asking questions at 4am?
The obvious suspect is the evaluator doing homework late at night because the day is full of meetings. That person exists. But if they drove the after-hours number, evenings would be the big bucket, and evenings are one of the smallest.
My theory for the large pre-dawn block: midnight to 9am Eastern is morning and afternoon in Europe, and the working day in India and much of Asia. A US team calls those hours "after hours". For the buyer typing the question, it is often the middle of a normal working day.
Our HubSpot notes do not store where a visitor was, so I can't prove the geography from this dataset. A second, larger dataset helps. Storylane's Modern Buying Trends 2026 report timed 8,211 website agent conversations across 19 companies on each buyer's own local clock. It found that 51.4% start outside the buyer's working hours, including 14.3% overnight (midnight to 6am) and 19.6% in the early morning (6am to 9am).
The two numbers measure different things. Our 51.6% is one website, timed on a US Eastern desk's clock. The report's 51.4% is 19 companies, timed on each buyer's clock. Together they say that about half of buyer conversations happen outside the buyer's own working day, and a big share start before the day begins. So the pre-dawn block on a US desk is partly other time zones and partly buyers reading before their own meetings start.
after-hours leads are mostly not late-night leads; they are leads from hours your team has never been awake for.
Are after-hours buyers less serious?
If the night shift were idle browsing, the conversations would be shorter and less qualified. They mostly aren't.
| Measure | Outside Eastern desk hours (n = 1,226) | Inside Eastern desk hours (n = 1,152) |
|---|---|---|
| Two or more messages in the conversation | 41.9% | 43.1% |
| Rated high intent | 44.1% | 49.7% |
| Qualified on our criteria | 28.5% | 32.2% |
| Ended with a meeting link shared | 9.4% | 13.3% |
Source: internal Storylane data (RepX website conversations), July 21 to November 11, 2026. Intent and qualification are the agent's own ratings, logged with each conversation.
Engagement is flat. Qualification is close. Storylane's Modern Buying Trends 2026 report found the same thing across its 19 companies: conversations that start outside working hours run just as long as the ones inside them. Out of the 720 conversations that qualified, 349 (48.5%) happened while an Eastern desk was closed.
The one gap worth reading is the meeting link, which overnight buyers end with less often. My read is that they are earlier in their work: they want the answer now and the call later. That matches what we saw in the weekend data, where 3 of 1,429 messages asked for a human, and the report's finding that requests for a human rise as conversations deepen: 25.4% of the longest website agent conversations end in a request for a human, against 1.6% of single-turn ones. Buyers use the agent to work out whether you are worth a call at all, then ask for the call. That is the case for treating an agent qualified lead as its own stage.
How long does an after-hours lead wait?
For each conversation that started outside Eastern desk hours, I measured how long it was until the next weekday 9am Eastern. That is the wait before anyone could begin to respond, before your speed to lead clock starts.
- Median wait: 5.2 hours.
- Mean wait: 9.8 hours (weekends pull it up).
- One in five (20.5%) would wait more than 12 hours.
Source: internal Storylane data (RepX website conversations), 1,226 conversations outside Monday to Friday, 9am to 6pm Eastern.
Five hours sounds survivable, but the buyer doesn't wait. They open the next vendor's site, which may already be open in another tab. A form that promises a reply "within one business day" is a reply that arrives after the buyer has formed a view, which is why contact forms suppress conversion more than most teams expect.
How do I cover inbound leads that arrive outside business hours?
There are three honest options, and each is right for someone.
- A form and a fast morning follow-up. Pick this if your inbound volume is small and nearly all of it comes from one time zone. Make the morning queue your SDRs' first job, sorted by arrival time, with a follow-up cadence that starts when the desk opens (see how to run inbound lead management), so overnight leads don't sit behind the day's new ones.
- A shift in another region. Pick this if you sell into Europe or Asia on purpose and need people who can run discovery calls in those hours. It covers the night shift with humans, at the cost of a second team. Our comparison of AI SDRs and human SDRs lays out the cost and quality trade.
- An agent that answers the question in full. Pick this if the night shift is a meaningful share of your conversations and the buyers in it want answers more than meetings, as our data shows for our site.
I'm biased on the third one, since we build it. RepX, Storylane's website agent, is built for teams whose buyers show up around the clock. It answers from your website, docs, decks and call recordings, pulls the relevant interactive demo into the conversation so the buyer can see the product (each demo has a plain-language display instruction that decides when it appears), and qualifies in real time on your ICP written in plain English. Playbooks set the starter questions, CTAs and qualification questions for each page, and a meeting goal can hand the buyer a calendar link from Chili Piper, RevenueHero or HubSpot Calendar at 4am without anyone awake.
Qualified leads and conversation summaries flow to HubSpot, Salesforce and Slack, so the rep who starts at 9am opens the night's conversations already knowing what each buyer asked. In HubSpot, each conversation is logged as a note with the full transcript, summary, topics and whether the buyer gave an email or booked a meeting. Those notes are the dataset behind this article, and they are what makes routing to the right owner in HubSpot work the next morning. At Hospitable, RepX booked 36 meetings and sourced 7 closed-won deals in its first month. RepX starts at $2,000 a month, and the price depends on website traffic.
Whatever the tool, the night shift needs an answer at the moment of the question, and a person in the morning who starts with context instead of a cold form fill.
How to measure your own night shift
You can run this on your own site this week. Export the timestamps of every chat and form submission for the last 90 days. Convert them to your sales desk's time zone, not the buyer's. Then count three things:
- The share before your desk opens, after it closes, and on weekends, as separate buckets.
- The share of your qualified or high-intent leads in each bucket.
- The median hours from each off-hours lead to your next opening time.
If the before-opening bucket is the big one, as it is for us, a later shift won't help and a better answer will. If you have run an agent for a while, our five learnings from 1,332 RepX conversations covers what else to look for in the logs.
How we counted
We pulled every RepX HubSpot note for a website conversation, read-only. 2,378 conversations after removing 5 duplicates, from July 21 to November 11, 2026; nearly 9 in 10 started on a storylane.io page. Timestamps are when the conversation was logged, converted from UTC to US Eastern and US Pacific. "Working hours" means Monday to Friday, 9am to 6pm on the sales desk's clock. We counted every conversation, including one-message ones; restricting to conversations with two or more messages (n = 1,011) gives 50.8% outside Eastern hours, so the pattern holds. This is one company's website, so treat it as a strong signal for B2B software, not a law.
To see what your buyers would get at 4am on your own site, book a demo with our team.