Ads bring in newcomers. Now you see who came back.
Every client in your base sits in her own segment and stage: normal, due to return, at risk, dormant, lost. The platform does the routine math every night; the marketer decides who to talk to and what to say.
You can see reach. Not the money from repeat visits.
Standard reports show bookings, not returns: a newcomer and a regular look exactly the same.
Messaging everyone annoys those already booked and doesn’t tell apart those who left for good.
Without cohorts you can’t show what matters: how many people became regulars and what that’s worth in money.
While no one tracks the interval between visits, the loss only shows up in the annual report.
Bringing clients back pays better than bringing new ones
A newcomer costs you ad spend. They pay off when they come back — and the second visit is exactly where salons lose the most.
Research and industry-platform data from other markets — a benchmark. You’ll see your own numbers in the reports.
Reports a marketer works with every week
Platform screens with demo data.
RFM and segments
| Segment | Clients | Not booked | What to do |
|---|---|---|---|
| Champions | 640 | 520 | No hard selling: say thank you, give early access to new services, ask for a review and a referral. Bot. |
| Loyal | 1 120 | 860 | Stay in touch, offer additional services and a loyalty program. Bot. |
| At risk | 380 | 350 | Personal message or call, a win-back offer. Bot, SMS for those not in the bot. |
| About to sleep | 720 | 690 | Remind them about you with a specific reason; a soft offer if they do not respond. Bot. |
| Hibernating | 1 450 | 1 420 | Mass reactivation via a cheap channel: news, a seasonal reason, an offer. Bot. |
| Lost | 2 310 | 2 300 | Free channel only (bot), rarely. Do not spend on SMS. |
Who to message — already decided
- Recency, frequency, monetary. Three numbers that are enough to predict a client’s value — as researchers at Wharton and London Business School have shown. [6]
- Advice for every segment. Who to thank, who to remind, who to call personally, and who not to spend SMS on.
- Stages by direction. The “due to return” threshold is based on your usual interval between visits, not a generic norm.
Segment builder
| Client | Channel | Last visit | Days ago | Visits | Segment |
|---|---|---|---|---|---|
| Client 1 | Bot | 18.07 | 96 | 6 | At risk |
| Client 2 | SMS | 21.07 | 93 | 4 | Need attention |
| Client 3 | Bot | 25.07 | 89 | 9 | Cannot lose them |
| Client 4 | Bot | 29.07 | 85 | 3 | About to sleep |
| Client 5 | SMS | 02.08 | 81 | 5 | Need attention |
A segment that never goes stale
- It saves the conditions, not a list. Tomorrow the segment will hold the people who are “due to return” tomorrow.
- SMS budget up front. How many people are SMS-only, how many parts the text takes and what it will cost.
- Exports with a log. Every export of the base is recorded: who, when, which filters.
Newcomer return
| Slice | Booked | Showed up | Came back | Twice or more | Days to return | Same specialist | Lost |
|---|---|---|---|---|---|---|---|
| Branch 1 | 168 | 146 | 8659% | 51 | 38 | 74% | 29 |
| Branch 2 | 131 | 112 | 5852% | 33 | 44 | 66% | 31 |
| Branch 3 | 113 | 98 | 5758% | 34 | 41 | 70% | 22 |
| Network | 412 | 356 | 20156% | 118 | 40 | 71% | 82 |
What marketing delivered — in people and visits
- From ad to second visit. How many newcomers booked, showed up, came back once, twice, three times — and after how many days.
- Where the person came back to. The same specialist, a different one, another branch — or nowhere.
- When people book. A heat map of booking hours and the share of online bookings show when to switch ads on.
Freed-up slots
| Branch | Cancellations | Hours | Refilled | Lost |
|---|---|---|---|---|
| Branch 1 | 18 | 27 | 21 | 6 |
| Branch 2 | 15 | 24 | 16 | 8 |
| Branch 3 | 13 | 20 | 15 | 5 |
A reason for a last-minute offer
- Where and when time frees up. By branch, specialist, direction and how many hours before the visit people cancel.
- Who refills it. See how many slots were resold and how many specialist hours went to waste.
Win-back sequence
The platform sets the task, a person makes the call
- On time. The manager gets a task on the very day the client is due to return — with the directions that are due.
- Closes itself. The client booked this direction — the task closes without the manager.
- Control group — on request. By default every client gets reminders. A network can set aside 10% with no outreach to see the real effect, not what would have happened anyway.
Try it in your business — the request takes a minute.
Try itSee month by month whether new clients become regulars
Clients are grouped by the month of their first visit. You then see how many come back after 1, 2 and 12 months, how much money they bring and where they drop off. The platform recalculates everything every night over the whole Altegio booking history.
| Cohort | Clients | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 10.2026 | 1 310 | |||||||||||
| 09.2026 | 1 180 | 27,9 | ||||||||||
| 08.2026 | 1 265 | 26,9 | 14,1 | |||||||||
| 07.2026 | 1 420 | 27,9 | 13,7 | 7,5 | ||||||||
| 06.2026 | 1 090 | 27,9 | 12,9 | 7,7 | 5,6 | |||||||
| 05.2026 | 1 150 | 18,8 | 13,1 | 7,7 | 5,2 | 4,7 | ||||||
| 04.2026 | 1 330 | 25,8 | 9,7 | 7,2 | 5,4 | 4,6 | 3,8 | |||||
| 03.2026 | 1 240 | 26,8 | 12,7 | 5,1 | 5,5 | 4,3 | 3,9 | 3,4 | ||||
| 02.2026 | 1 010 | 26,0 | 12,2 | 7,4 | 3,7 | 4,3 | 4,0 | 3,2 | 2,9 | |||
| 01.2026 | 1 120 | 24,5 | 12,6 | 7,2 | 5,0 | 3,2 | 3,8 | 3,1 | 3,0 | 2,6 | ||
| 12.2025 | 1 385 | 24,8 | 12,6 | 6,8 | 5,2 | 4,3 | 2,6 | 3,2 | 2,8 | 2,6 | 2,6 | |
| 11.2025 | 1 205 | 25,5 | 11,8 | 6,9 | 5,1 | 4,1 | 3,8 | 2,3 | 2,7 | 2,7 | 2,5 | 2,4 |
| Average | 25,7 | 12,5 | 7,0 | 5,1 | 4,2 | 3,6 | 3,0 | 2,8 | 2,6 | 2,6 | 2,4 |
How to read
A row is the new clients of one month, a column is how many months have passed. A cell shows the share of the cohort that came back exactly then.
Newer rows higher than older ones in the same column — you have got better at bringing clients back. A dip along the diagonal is a calendar event: season, prices, a specialist.
What to do
Find the month after which cohorts got better, understand what changed then, and keep doing it.
Hover over a cell: the row, the column and, in orange, the same calendar month in other cohorts light up. Here you can see a seasonal dip 5 months ago.
How to read
A drop in the first months is normal: some clients come only once. What matters is that the curve levels off into a flat “shelf” instead of falling to zero.
What to do
Raise the shelf: a treatment plan and booking the next visit during the first one.
Of 1,000 new clients — how many reached each next visit
How to read
Most clients are lost between the 1st and 2nd visit. Those who come back stay better after that.
What to do
The “Did not return after the first visit” list goes to an administrator: call, ask how the visit went, offer a convenient time.
Demo data. You will see your own value by cohort in your account.
How to read
Payback is the time it takes for the profit from a client to cover the cost of acquiring them. So compare the cost of acquisition (ad spend ÷ new clients) with margin, not revenue.
What to do
If payback takes longer than a year, first learn to bring new clients back, then increase advertising.
of new clients return for a 2nd visit at an average salon; 70% at the best ones
Boulevard, over 11 million appointments, 2023 ↗ +25–95%profit from a 5% increase in customer retention
Bain & Company · Harvard Business Review, 2014 ↗ Newer is bettera healthy sign: retention of each cohort levels off over time, and newer cohorts retain better than older ones
a16z, 16 More Startup Metrics ↗ Shelfif the retention curve drops at first and then flattens out instead of going to zero, you have a core of clients who stay
Amplitude ↗How the platform counts client stages ↗
Platform automatically
- Recalculates cohorts every night over the whole booking history.
- Does not show months that are not over yet, so young cohorts do not look worse than they are.
- Prepares a list of those who did not return after the first visit.
People decisions
- The marketer looks for the changes that lifted cohorts and puts advertising where new clients come back from.
- The administrator calls those who did not return.
- The owner compares branches and specialists — to learn from the best, not to punish.
Try it in your business — the request takes a minute.
Try itWho your revenue rests on — and why it changed
Three charts from cohort analysis: each month’s revenue in layers by year of first visit, the change in revenue broken down into its causes, and how many new clients are still coming a year later. The platform recalculates them every night over the whole booking history.
share of clients who first came in the last 12 months: 37%
share of clients who first came in the last 12 months: 83%
How to read
The old layers barely thin out: clients who came years ago still bring in most of the revenue. New clients stack on top — the business is growing.
Every layer thins out fast and fresh newcomers carry the revenue — the red line is high. Stop advertising, and within a few months revenue drops too.
What to do
If the red line is rising while the old layers melt away, first learn to bring clients back, then spend more on advertising.
Switch between two demo networks with the same revenue — the difference is obvious at once. Hover over a bar to see each layer’s share.
The scale does not start at zero, so the steps are visible.
Move the sliders — the steps and the total recalculate.
How to read
Revenue can grow while clients come less often. The breakdown shows exactly which lever moved: a price rise shows up in “revenue per visit”, people you did not bring back — in “how many regulars came”. The steps always add up to the change in revenue.
What to do
Each step has its owner: new clients — the marketer; “how many came” and “how often” — administrators and reminders; “per visit” — specialists and services.
months after the first visit
| Slice | New clients | 1 mo. | 3 mo. | 6 mo. | 12 mo. | 24 mo. |
|---|---|---|---|---|---|---|
| Whole network | 1 100 | 38% | 32% | 26% | 20% | 17% |
| Branch A | 620 | 42% | 36% | 30% | 24% | 22% |
| Branch B | 480 | 33% | 26% | 20% | 14% | 12% |
How to read
The steep drop in the first month is the clients who came only once. What matters next is the height at which the curve levels off. Young cohorts count too: each client is followed up to the month by which they could have come back.
What to do
Compare branches and specialists: where the curve is higher, find out what they do differently there and pass it on to the others.
at a retailer with weak retention, new customers bring about 90% of monthly revenue; at one with strong retention, almost 80% comes from returning customers
Custora · Amperity, 2020 ↗ C3analysts call the cohort revenue chart the most important customer-related disclosure investors should ask for
Theta Equity Partners, 2020 ↗ CBCVa company’s value is estimated from the bottom up: how many customers it acquires, how many it loses and how much they spend
McCarthy & Fader · Harvard Business Review, 2020 ↗ 1958the Kaplan–Meier method accounts for those whose history is still going on: they count while they are visible and do not skew the result
Bland & Altman · BMJ, 1998 ↗Platform automatically
- Recalculates the cake, the breakdown and survival every night over the whole booking history.
- Breaks down the revenue change exactly: the steps add up to the difference.
- Opens each chart as a separate report with filters by branch, specialist, category and source.
People decisions
- The owner sees whether the business rests on regulars and decides where to invest.
- The CFO explains the change in revenue by its causes, not by “the season”.
- The marketer compares branches by survival and looks for what works.
The right word to the right person — at the right time
The rules are built in so clients don’t get annoyed.
We leave booked clients alone
Anyone already booked ahead is set aside: they don’t need a return reminder.
Lost clients — bot only
For people long gone, RFM advises against flooding them with SMS: a rare free touch, and only where the person subscribed themselves.
Effect vs. control
Leave part of the segment without outreach and compare: you’ll see what the campaign did, not the calendar.
Marketing brought them in. Did the front desk book them?
Whether a newcomer comes back is often decided at the first visit: did the front desk rebook, confirm and get them to the visit. That’s what Front desk analytics measures.
The plan for marketers
Statuses come from our work plan and change along with it.
- RFM, stages and saved segments
- Newcomer return by branch and specialist
- Cohort analysis: retention triangle, client value, visit ladder
- New clients: who booked and who rebooked
- Freed-up slots
- Segment campaigns with a control group
- SMS: AlphaSMS, TurboSMS, Messaggio
- Win-back sequences
- Recommended date — sent to the client in the bot
- Client tags
- Full NPS cycle and reviews
- End-to-end ad analytics: from click to LTV
What winning back dormant clients is worth
Take dormant and lost clients from RFM, set a realistic share who will come back and how many times a year they’ll visit.
- Clients returned—
- Visits per year—
This is revenue, not profit, and the estimate is approximate: some people would have come back anyway. A control group, if the network turns it on, will show it more precisely.
Try it in your business — the request takes a minute.
Try itA week where you can see the money
Five short actions that keep your base alive.
Segments of the week
“Due to return” and “At risk” — new people showed up over the weekend.
Freed-up slots
Where specialists’ time goes to waste — that’s where the last-minute offer goes.
Cohort of the month
Which newcomers came back and to whom — what works in your ads.
Report to the owner
Returns and money, not views.
Ad adjustments
Booking hours, online booking, directions where newcomers don’t come back.
A direct-marketing method: whoever bought recently, often and for more is the most likely to respond.
Source [5]Recency, frequency and monetary value are enough to forecast a client’s value.
Source [6]Value = dream outcome × perceived likelihood of achievement ÷ (time delay × effort and sacrifice). Alex Hormozi, $100M Offers.
Source [7]There are only four ways to get clients; the cheapest is your warm audience — your past clients. Alex Hormozi, $100M Leads.
Source [8]What this page is based on
Every source is linked — open it and see for yourself.
- Harvard Business Review — The Value of Keeping the Right Customers (2014) — acquiring a new client costs 5–25 times more than keeping one.
- Boulevard — Top-performing salons retain 56% more first-time visitors (2023) — more than half of newcomers don’t come back; at the best salons 70% return; online booking — 78% vs. 39%.
- Kitomba — Rebooking rates for the hair and beauty industry (2021) — average beauty salon rebooking rate is 43–47%, the best — 80% and up.
- Bain & Company — Frederick F. Reichheld, Prescription for cutting costs (2001) — in financial services, +5% retention yields more than +25% profit.
- Arthur M. Hughes — Strategic Database Marketing (McGraw-Hill, 2000) — RFM as a tool for selling to existing clients.
- Fader, Hardie, Lee — RFM and CLV: Using Iso-Value Curves for Customer Base Analysis (Journal of Marketing Research, 2005) — RFM and customer value: three numbers are enough for a forecast.
- Peter Yang — The value equation for irresistible products (on Alex Hormozi, $100M Offers) — the value equation.
- Shortform — $100M Leads by Alex Hormozi — four ways to acquire clients.
We first test these approaches in practice with the salons we work with. We don’t name them or share their numbers — it is their data, and we protect their confidentiality. That is why we cite open sources: you can open and check each one yourself.
Pages for the other roles in the salon
Owner
The bookings are in alteg.io, but where the revenue goes is invisible.
Where money leaks and how much of it you can win back — with a calculator on your own numbers.
For ownersFront desk
Who confirms visits and rebooks clients, and who doesn’t, gets settled in arguments.
20 metrics for every shift, daily in Telegram.
Front desk analyticsSpecialist — stylist, doctor
Client protocols and photos live in chats and on the specialist’s phone.
A cabinet that opens right from Telegram: client card, a “same as last time” protocol, photos never stay on the phone.
Specialist cabinetAccountant
Three statements, Excel and an alteg.io cash desk that “almost” matches.
Daily bank and cash desk reconciliation — the accountant sees only the mismatches.
For accountantsCFO Planned
The report shows a profit, but there’s less cash in the account.
Cash flow, profit and loss, cost of acquiring — by branch and network.
For CFOsFor those already working with the platform: what works, what we are building now and what is planned — and you can vote on what matters most.
CRM: where we’re headingWe’ll show it on your base
Send a request — we’ll connect your salon and walk through segments and newcomer return on your own data.
Or write to us directly:
founder@alteg-tech.com