Luna Studios Research · Model calculation

The Answer Dividend.

What happens, arithmetically, when a website can answer? A model calculation based on our own measurement of 819,686 business websites and the established research on response time — computed for four typical businesses, with all assumptions disclosed.

Model calculation, not a customer study All assumptions disclosed Conservatively computed

The starting point — measured, not guessed

Four numbers that precede any calculation.

99,6 %of business websites in the DACH region have no channel for an immediate answer — own measurement, 2026

70 %of the week the typical website is on its own: median availability 8 am – 6 pm, 50 of 168 weekly hours — own measurement, 2026

83 %of people expect to interact with someone immediately when they reach out — independent consumer research

23 %of online inquiries to 2,241 audited companies were NEVER answered — Harvard Business Review

Between what visitors expect and what websites deliver lies a gap. This calculation puts a number on what is lost in that gap — and what an answering website brings back.

The model

Five steps, one chain.

Every scenario follows the same calculation. Each step is either our own measurement, a documented third-party source, or a conservative assumption — the origin is stated at every number, and all assumptions are tabulated at the end.

1Monthly website visits of the model business (assumption, conservative)

2Share of visits with a concrete request — appointment, price, availability (assumption)

3Share of requests outside attended hours (derived from own measurement)

4Share that moves on without an immediate answer instead of waiting (assumption)

5Share of those an answering website serves immediately (assumption: 80 %)

The result of the chain is recovered requests — conversations that end wordlessly today. The revenue line below is an illustration using typical industry averages; it shows the order of magnitude, not a guarantee.

Scenario 01

The restaurant.

Model business: restaurant, busy, website with menu

Industry Speaking Index: 0.03 — one in 3,553 can answer

2.000website visits per monthassumption

300of them with a concrete request: reservation, group, allergens, opening hours15 % — assumption

150of those outside attended hours — during lunch service, after the kitchen closes50 % — derived

90move on to the next restaurant without an immediate answer60 % — assumption

72are recovered monthly by an answering website80 % — assumption

≈ €26,000 / year

Illustration: if one third of recovered requests leads to a visit (24/month) at an average table revenue of €90 → €2,160/month. 864 requests a year that end wordlessly today.

Scenario 02

The dental practice.

Model business: dental practice, two practitioners, office hours Mon–Fri

Own measurement: 0 of 5,297 German dental practices can answer

1.200website visits per monthassumption

120of them with a concrete request: appointment, pain, cost question10 % — assumption

72of those outside office hours — evenings, Wednesday afternoons, weekends60 % — derived

36call the next practice instead of waiting50 % — assumption

29are recovered monthly by an answering website80 % — assumption

≈ €21,600 / year

Illustration: if a quarter of recovered requests becomes an appointment (7/month) at an average first treatment of €250 → €1,800/month — excluding the lifetime value of a new patient, which is a multiple of that. And: the answering practice would currently be the only one in the country.

Scenario 03

The trades business.

Model business: trades company, 8 employees, on job sites during the day

Industry Speaking Index: 0.02 — the lowest of all twelve groups

600website visits per monthassumption

90of them with a concrete request: quote, emergency, availability15 % — assumption

59of those outside reachable hours — customers research after work65 % — derived

29ask the next company without an immediate answer50 % — assumption

23are recovered monthly by an answering website80 % — assumption

≈ €63,000 / year

Illustration: if 15 % of recovered requests becomes a job (3.5/month) at an average job value of €1,500 → €5,250/month. In the trades every single recovered inquiry counts — and 23 % of all online inquiries are never answered today.

Scenario 04

The online shop.

Model business: online shop, established, 50,000 visits per month

Retail Speaking Index: 0.07 — one in 1,537 can answer

50.000website visits per monthassumption

2.000of them with a concrete request: size, shipping, returns, product question4 % — assumption

1.200of those outside service hours — people shop in the evening and on weekends60 % — derived

720abandon without an answer or order from the next shop60 % — assumption

576are recovered monthly by an answering website80 % — assumption

≈ €110,000 / year

Illustration: if a fifth of recovered requests leads to a purchase (115/month) at an average basket of €80 → €9,216/month. At scale, the dividend grows linearly: every unanswered question at checkout is an abandoned purchase — and here it is asked two thousand times a month.

The second dividend

Speed matters even when someone is there.

The chain above only counts requests outside attended hours. Response-time research shows that speed decides outcomes during opening hours too:

100×more often a lead is reached when the response comes within 5 instead of 30 minutes — MIT / InsideSales, 15,000+ inquiries

7×more often an inquiry is qualified when answered within one hour — Harvard Business Review, 2,241 companies

Secondsis the response time of an answering website — before any callback, before any email, around the clock

An answering website moves every single first contact from “sometime” to “now” — and that is exactly where the documented multipliers live.

The full picture

Four businesses, one calculation.

Model business	Requests/month	lost today	recovered	Answer dividend/year*
Restaurant	300	90	72	≈ €26,000
Dental practice	120	36	29	≈ €21,600
Trades business	90	29	23	≈ €63,000
Online shop	2.000	720	576	≈ €110,000

* Revenue illustration using typical industry averages and conservative close rates; derivation per scenario, assumptions below.

The core statement of the model

€21,600 – €110,000

arithmetical answer dividend per business per year

Not through more visitors. Not through more advertising. Simply because the website answers what visitors already want to ask it.

The model calculator

Compute your own answer dividend.

Every lever of the model, freely adjustable. The defaults match the scenarios' conservative assumptions — every number can be set to your business.

Website visits per month 2.000

Visits with a concrete request 15 %

Requests outside attended hours 50 %

Move on without an immediate answer 60 %

Recovery rate of the answering website 80 %

Requests that convert 33 %

Average value per conversion €90

72recovered requests per month — conversations that end wordlessly today

≈ €26,000arithmetical answer dividend per year (€2,160 per month)

Model calculation following the chain described above — an illustration of the order of magnitude, not a revenue guarantee. Defaults: the scenarios' conservative assumptions.

Assumptions & sources

Every number has an origin.

Building block	Value	Origin
No immediate-answer channel (DACH)	99.6 %	Own measurement, “The Talking Web 2026”, 819,686 websites
Unattended weekly hours	70 % (118 of 168)	Own measurement: median availability 8 am – 6 pm on weekdays
Speaking Index per industry	0.02 – 0.07	Own measurement, industry excerpts 2026
Expectation of immediate interaction	83 %	Independent consumer research (evidence: /en/research/theses)
Never-answered inquiries	23 %	Harvard Business Review, audit of 2,241 companies
Response-time multipliers	100× / 21× and 7× / 60×	MIT/InsideSales and Harvard Business Review
Monthly website visits	600 – 50,000	Assumption per business type, deliberately at the lower end of typical
Share of visits with a request	4 – 15 %	Assumption, conservative
Requests outside attended hours	50 – 65 %	Derived from 70 % unattended weekly time, rounded down conservatively
Move on without immediate answer	50 – 60 %	Assumption, supported by the 83 % expectation and response-time research
Recovery rate of answering website	80 %	Assumption (not every request can be resolved automatically)
Close rates of the illustration	15 – 33 % (shop: 20 %)	Assumption, conservative; industry-typical averages

Method & limits

What this is: a model calculation — the documented measurements of the research series, combined with documented third-party sources and disclosed, conservative assumptions. Not a measurement of customer installations, not a revenue guarantee.
Why conservative: every assumption is deliberately chosen at the lower end; the lifetime value of returning customers, referrals, and the speed dividend during opening hours are not included.
What follows: once sufficient measurement data from real installations is available (with customer consent, anonymized), the assumptions will be replaced step by step with measured values — with the same openness as the rest of the series.
Usage: freely quotable as “model calculation by Luna Studios, lunastudios.ai/research” — please always label it as a model calculation.

And now?

Check your own business.

Where does your website stand?

Luna, the assistant on lunastudios.ai, looks at your website in real time when you tell her your domain — and tells you which of the requests computed here she could already answer today.

Check your website → The research series →