GuestKeepr

For marketing

Where your guests come from — and where there is more revenue

When the host seats a guest, they mark where the guest came from. From these marks the dashboard works out what a new guest costs in each channel, shows on which days and at which hours the room stands empty, and helps you check a promotion with numbers. Both calculations are here, on a made-up venue; put your own numbers into the calculators.

Part 1 · Where guests come from and what a new one costs

What the statistics give a marketer

When the host seats a guest, they mark where the guest came from. You enter your advertising spend month by month. The dashboard works out what a new guest costs in each channel.

An example with made-up data. Your own numbers will be in your dashboard.

Where the guests came from

Over a month, from the host’s marks at seating.

  • Regular guests64
  • Google38
  • 2GIS31
  • Word of mouth27
  • TikTok22
  • Walked past18
  • Promoter12

Peak hours and empty hours

Seatings by the hour over a month. The early evening is nearly empty: it asks for an offer or an event, not one more ad for 23:00.

518:00 — 5, nearly empty
919:00 — 9, nearly empty
1820:00 — 18
3421:00 — 34
4722:00 — 47
5123:00 — 51, peak
3800:00 — 38
2101:00 — 21
802:00 — 8, nearly empty

30%

Share of regulars

If it falls, inviting new guests isn’t enough: it’s worth finding out why the old ones don’t come back. Regulars aren’t counted in the cost of a new guest.

Reallocate the budget

The total stays the same; only where it goes changes. Move the sliders and watch what a new guest costs.

  • Spend 400 € · new guests 38 · 10.5 €
  • Spend 250 € · new guests 31 · 8.1 €
  • Spend 300 € · new guests 22 · 13.6 €
  • Spend 450 € · new guests 12 · 37.5 €

New guests: 103. Cost of a new guest: 13.6 €, was 13.6 €.

Monthly budget

1,400€

New guests

103

was: 103

Cost of a new guest

13.6€

was: 13.6 €

How we counted

A channel’s new guests = last month’s guests × (new spend ÷ last spend) to the power of 0.7. Cost of a new guest = spend ÷ new guests. It’s a toy model: each further guest from a channel costs a little more than the one before, so putting everything into one channel isn’t the best answer. Your dashboard will show the real answer in a week.

A marketer’s week

  1. Saw the numbersThe Marketing section shows which channel brings new guests cheaper and which costs more.
  2. Changed the campaignMoved part of the money out of the expensive channel, or changed the ad itself.
  3. Checked a week laterOpened the same report for the new period. Not cheaper? Next hypothesis.
The Marketing section of the dashboard: the cost of a new guest, spend and channels
This is how it looks in the dashboard: spend by channel and the cost of a new guest.

Part 2 · Where there is more revenue

The statistics show where there is more to earn

The dashboard counts how many guests come per seating, how long they stay at the table, how many guests each day brings and at which hours there are the most. Each of these numbers can be nudged up a little. A little every day adds up to real money over a month.

An example with made-up data: a month of an imaginary restaurant. Put in your own numbers.

What the numbers show

A restaurant’s month

  • 2.3

    guests per seating on average

    Most come as a couple

  • Sun and Mon

    the weakest days

    32% and 44% fewer guests than on an average day

  • 15:00–18:00

    the emptiest hours

    3.5 times fewer guests than at 20:00

When the room is empty

Guests by day of the week and hour of seating, on an average day

Guests by day of the week and hour of seating, on an average day
12:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:00
Monday57522357752
Tuesday7973247111274
Wednesday71074348121495
Thursday710844591416127
Friday71185471219211610
Saturday81097791421231912
Sunday810732258852

The decision

Example offer

Sunday and Monday · 15:00–18:00

Come as three — dessert for all, on the house

The gift and the terms are yours

The offer works on two numbers at once: it brings guests in at hours when the tables stand empty anyway, and it makes the parties bigger. A dessert costs less than an empty table.

Add the offer to the venue profile, under Promotions, and the bot will tell a guest about it when they ask. Where else to announce it is up to you.

What it is worth

Currency

Your numbers

In the chosen currency

A month usually has eight or nine Sundays and Mondays

On one such day

In the dashboard: “guests per seating on average”

Your estimate

Nobody knows in advance what an offer will bring. The starting values are cautious: put in your own, and check them against the dashboard a month later.

On one such day

Some couples will come as three

Per guest, as a share of that guest’s spend. Generously: as if every guest in those hours got the gift

Extra revenue a month

+419€

  • Bigger parties · +12 guests+240 €
  • New seatings · +22 guests+448 €
  • Gifts−269 €

A year, on your assumptions

+5,030€

Extra revenue a month: 419 €. A year, on your assumptions: 5,030 €.

How we counted
  1. Now: 8 days × 3 seatings × 2.3 guests = 55.2 guests in those hours a month
  2. With the offer: 8 days × 4 seatings × 2.8 guests = 89.6 guests
  3. Extra: (89.6 − 55.2) guests × 20 € = 688 €
  4. Gifts: 89.6 guests × 15% of 20 € = 269 €
  5. In total: 688 € − 269 € = 419 € a month, 5,030 € a year

This is revenue, not profit: only the gifts are deducted, not the cost of the food. It is a calculation from your inputs, not a promise: an offer may not work, which is why it is checked against the numbers.

More numbers that are worth money

The same move: find a number in the dashboard, decide, count. The spend per guest and the currency come from the calculation above.

Restaurant or club

The table is free sooner than the bot thinks

In the dashboard
Guests stay 2 h 10 min on average. Yet the bot treats a table as taken for three hours after seating: that is the default until you change it.
The decision
Press “Set 2 h 10 min” in Statistics. From then on the bot treats a table booked for 18:00 as free from 20:10, not 21:00, and can take a booking for 20:30 on it.

If the bot never turns guests away anyway, there is nothing to gain

+368€a month

8 evenings × 1 bookings × 2.3 guests × 20 €

368 € a month

The Statistics section of the dashboard: guests stay 2 h 10 min on average, the bot treats a table as taken for 3 hours, a button sets 2 h 10 min
This is how it looks in the dashboard (its interface is in Russian): the average time at the table and the button that hands it to the bot.

Club

Empty until midnight

In the dashboard
In Peak hours the top places go to the hours after midnight, and the early evening does not make the six busiest hours.
The decision
Early entry: parties who arrive before 23:30 get a cocktail from the bar. Add it to Promotions in the profile, and the bot will tell a guest about it when they ask.

Those who would not have come without the offer. Don’t count those who would have come anyway, only later

As a share of one guest’s spend

+864€a month

8 nights × 6 guests × 20 € − cocktails 96 €

864 € a month

Any venue

A new guest is cheaper in another channel

The host marks where each guest came from, you enter your advertising spend, and the dashboard works out what a new guest costs in each channel. Move part of the money to where a guest is cheaper: the same money, more guests.

Try it above

Cost of a new guest by channel, the example above

  • Google10.5 €
  • 2GIS8.1 €
  • TikTok13.6 €
  • Promoter37.5 €

A month later, check

An offer is a guess until it has been checked. The dashboard shows whether it worked.

  1. Saw the numbers2.3 guests per seating, Sunday and Monday the weakest days, nobody from 15:00 to 18:00.
  2. Ran the offerSunday and Monday, 15:00–18:00, a free dessert for every guest. Added it to the venue profile, so that the bot tells guests about it.
  3. ComparedThe next morning Overview compares yesterday with the same day a week earlier. A month later Statistics shows, next to guests, seatings and “guests per seating on average”, what they were in the previous period. If it didn’t work, change the terms.

Every sum in this section is a calculation from your inputs, not a forecast. You set the average spend yourself: GuestKeepr is not connected to your till and does not see your revenue.

Your own numbers will be in your dashboard

The host marks the source when seating a guest, and you enter your advertising spend. The dashboard does the rest: by day, by hour and by channel.

Get in touch

How the bot answers guests — on the home page