Your site knows who looked. Your calendar knows who booked.
Your register knows who paid. 314 knows it's one person, and brings them back.
Built and advised by a team from
A dozen tools. Not one of them notices when a customer stops coming.
Site, bookings, payments and ads land in one profile per customer. Anything a customer does, a visit, a booking, a payment, can set off a message.
314 finds who drifted, who read the prices and left, what stopped working. Each finding comes with a plan to fix it.
Email, SMS, popups and landing pages work as one sequence on one profile. 314 writes it. You approve it. From then on it reaches every customer who does the same thing.
The message, the click, the booking, the payment. 314 judges each sale the way a marketer would: did the message cause it, or not.
What worked and what didn't goes back into the analysis. Weak sequences get reworked, new ones proposed. Other tools stop at send. 314 keeps going.
An AI agent works inside 314 the way a marketer would on your team. You stay the one who decides.
Every decision opens with what it found and why, in plain words, and the test it set for itself.
What it created, what it edited, what it switched on.
What reaches your customers waits for your yes.
Two switches set how much it does without you. Every decision still goes on the record, signed by whoever made it.
Before a message reaches you, a second pass reads it for made-up prices, wrong facts and a missing unsubscribe.
Connect 314 to Claude or ChatGPT. Ask about your customers, plan a campaign, have it built. Every draft waits for your yes.
Connect in two minutesHow did last week compare to the week before, and why?
Recall the client’s rules and facts ›
Read results and revenue ›
Query the project’s data with SQL ›
Query the project’s data with SQL ›
Found the main driver: returning clients stopped booking online. Checking one more thing: whether that was a gap on specific days or spread across the week.
Query the project’s data with SQL ›
Last week (Sep 14 to 20) was weaker than the week before. Web traffic was flat, so fewer visitors booked. It wasn’t a traffic drop.
| Sep 7 to 13 | Sep 14 to 20 | |
|---|---|---|
| Sessions | 296 | 273 |
| Bookings | 55 | 38 (−31%) |
| Booking rate | 18.6% | 15.1% |
| Cancellations | 21 | 17 |
Why:
These are small numbers, and the reasons are my inference. What I measured is where the drop happened, not what caused it.
Next:
Plan a Thanksgiving campaign for our regulars and build it: the emails and the sequence.
Recall the client’s rules and facts ›
Read strategies ›
Read segments ›
Discover events, fields and values ›
Query the project’s data with SQL ›
Audience is set: 116 regulars (2+ paid visits, one in the last 90 days). Creating the segment, then the three emails.
Create or change segments ›
Read a playbook ›
Generate an image for an asset ›
Generate an image for an asset ›
Both covers are done. Writing the emails now.
Create or change emails ›
Create or change emails ›
Create or change emails ›
The emails are drafted. Wiring them into the sequence.
Create or change strategies ›
Preview a strategy ›
Preview an email ›
The Thanksgiving campaign is built: an audience of regulars, three emails and the sequence that sends them. Everything is a draft, and nothing sends until you approve it, before Tuesday, Nov 3, when the first email goes out.
Why it’s built this way. Your rule is that a regular rebooking on schedule isn’t money we brought in. So the booking emails are a courtesy, and the gift card email is where new money comes from.
| When | Who | |
|---|---|---|
| Tue Nov 3 | Regulars not booked in 14 days | “Ready for the table”: book before the holiday, bring someone new and you each get $15 |
| Tue Nov 17 | Only those still not booked | A short plain letter: Thanksgiving is next Thursday |
| Fri Nov 27 | Everyone in the flow | “Someone at your table”: a gift card in any amount that never expires |
What it has to beat: only 8 gift cards sold since April, and Thanksgiving week has 1 booking so far.
Before you approve:
Scheduled: on Nov 5 I check unsubscribes from the first email and recommend pausing Nov 17 if more than 2% leave. On Dec 12 I measure gift card sales against the baseline and say whether to repeat it for Christmas.
Find the clients who haven’t been back in 60 days and write them a win-back email.
Recall the client’s rules and facts ›
Discover events, fields and values ›
Read segments ›
Read emails ›
Query the project’s data with SQL ›
Query the project’s data with SQL ›
Query the project’s data with SQL ›
Before writing: who hasn’t been back, minus who is already in a flow or can’t be emailed.
Read a playbook ›
Discover events, fields and values ›
Create or change segments ›
Segment saved: 74 people. Now the letter.
Create or change emails ›
Preview a segment ›
Preview an email ›
I built the segment and the email. Both are drafts, so nothing goes out yet.
Who’s in it: 74 people. Paying clients with an email and a receipt in the last year. None has paid or booked in 60+ days, and we haven’t emailed any of them in 45 days.
Why a new letter. 177 clients haven’t been back in 60 days. 8 are on the do-not-contact list and 2 already have a booking. Of the 167 left, 92 are in the Reactivation flows right now. The other 74 went through Reactivation in August and didn’t come back, mostly one manicure and never again. “We miss you” has been tried on them, so this letter takes a different angle.
The email is a plain letter with no discount and one “Book a time” button, in two variants split 50/50:
The copy check passed both. I didn’t write “Monday” anywhere: notes from an earlier session say the studio opens only about half of Mondays.
Should I set up a one-time send to this segment? It will come to you as a decision to approve.
People read our prices and leave without booking. Build a popup to catch them.
Recall the client’s rules and facts ›
Read a playbook ›
Read popups ›
Query the project’s data with SQL ›
Query the project’s data with SQL ›
Discover events, fields and values ›
The data’s in. Next: where the prices sit on the live page and what the booking link is. Then the popup.
Read strategies ›
Query the project’s data with SQL ›
Read emails ›
Generate an image for an asset ›
Create or change popups ›
Create or change strategies ›
Preview a popup ›
I built the popup and the flow that shows it. Both are drafts, and nothing reaches visitors until each is approved.
What I found. There’s no pricing page: the prices sit in “Signature Care”, a quarter of the way down the homepage. In 60 days about 726 sessions reached them. Only 12.7% of those clicked Book and 1.1% converted, against 4.2% across all sessions. Most who clicked Book were on phones, which shaped the trigger.
The popup has two variants, 50/50, each with a Book button and a small “not today? email me the menu” field:
When it shows: homepage visitors who never paid and haven’t booked in 60 days, once they scroll past the prices, then on exit or after 15 seconds idle. Exit intent never fires on phones, so the idle rule covers mobile. At most twice, a week apart, and never before the welcome popup.
One thing to check: the $15 offer is live on the site, but 0 of 183 first-visit receipts carried a discount. Confirm it’s given at checkout, or the popup promises something the till doesn’t deliver.
Once it’s live I’ll judge it on Book clicks and bookings within 7 days, A against B, against the 12.7% baseline.
How did last week compare to the week before, and why?
UsingUsed 314 ›
Found the main driver: returning clients stopped booking online. Checking one more thing: whether that was a gap on specific days or spread across the week.
UsingUsed 314 ›
Last week (Sep 14 to 20) was weaker than the week before. Web traffic was flat, so fewer visitors booked. It wasn’t a traffic drop.
| Sep 7 to 13 | Sep 14 to 20 | |
|---|---|---|
| Sessions | 296 | 273 |
| Bookings | 55 | 38 (−31%) |
| Booking rate | 18.6% | 15.1% |
| Cancellations | 21 | 17 |
Why:
These are small numbers, and the reasons are my inference. What I measured is where the drop happened, not what caused it.
Next:
Plan a Thanksgiving campaign for our regulars and build it: the emails and the sequence.
UsingUsed 314 ›
Audience is set: 116 regulars (2+ paid visits, one in the last 90 days). Creating the segment, then the three emails.
UsingUsed 314 ›
Both covers are done. Writing the emails now.
UsingUsed 314 ›
The emails are drafted. Wiring them into the sequence.
UsingUsed 314 ›
UsingUsed 314 ›
The Thanksgiving campaign is built: an audience of regulars, three emails and the sequence that sends them. Everything is a draft, and nothing sends until you approve it, before Tuesday, Nov 3, when the first email goes out.
Why it’s built this way. Your rule is that a regular rebooking on schedule isn’t money we brought in. So the booking emails are a courtesy, and the gift card email is where new money comes from.
| When | Who | |
|---|---|---|
| Tue Nov 3 | Regulars not booked in 14 days | “Ready for the table”: book before the holiday, bring someone new and you each get $15 |
| Tue Nov 17 | Only those still not booked | A short plain letter: Thanksgiving is next Thursday |
| Fri Nov 27 | Everyone in the flow | “Someone at your table”: a gift card in any amount that never expires |
What it has to beat: only 8 gift cards sold since April, and Thanksgiving week has 1 booking so far.
Before you approve:
Scheduled: on Nov 5 I check unsubscribes from the first email and recommend pausing Nov 17 if more than 2% leave. On Dec 12 I measure gift card sales against the baseline and say whether to repeat it for Christmas.
Find the clients who haven’t been back in 60 days and write them a win-back email.
UsingUsed 314 ›
Before writing: who hasn’t been back, minus who is already in a flow or can’t be emailed.
UsingUsed 314 ›
Segment saved: 74 people. Now the letter.
UsingUsed 314 ›
UsingUsed 314 ›
I built the segment and the email. Both are drafts, so nothing goes out yet.
Who’s in it: 74 people. Paying clients with an email and a receipt in the last year. None has paid or booked in 60+ days, and we haven’t emailed any of them in 45 days.
Why a new letter. 177 clients haven’t been back in 60 days. 8 are on the do-not-contact list and 2 already have a booking. Of the 167 left, 92 are in the Reactivation flows right now. The other 74 went through Reactivation in August and didn’t come back, mostly one manicure and never again. “We miss you” has been tried on them, so this letter takes a different angle.
The email is a plain letter with no discount and one “Book a time” button, in two variants split 50/50:
The copy check passed both. I didn’t write “Monday” anywhere: notes from an earlier session say the studio opens only about half of Mondays.
Should I set up a one-time send to this segment? It will come to you as a decision to approve.
People read our prices and leave without booking. Build a popup to catch them.
UsingUsed 314 ›
The data’s in. Next: where the prices sit on the live page and what the booking link is. Then the popup.
UsingUsed 314 ›
I built the popup and the flow that shows it. Both are drafts, and nothing reaches visitors until each is approved.
What I found. There’s no pricing page: the prices sit in “Signature Care”, a quarter of the way down the homepage. In 60 days about 726 sessions reached them. Only 12.7% of those clicked Book and 1.1% converted, against 4.2% across all sessions. Most who clicked Book were on phones, which shaped the trigger.
The popup has two variants, 50/50, each with a Book button and a small “not today? email me the menu” field:
When it shows: homepage visitors who never paid and haven’t booked in 60 days, once they scroll past the prices, then on exit or after 15 seconds idle. Exit intent never fires on phones, so the idle rule covers mobile. At most twice, a week apart, and never before the welcome popup.
One thing to check: the $15 offer is live on the site, but 0 of 183 first-visit receipts carried a discount. Confirm it’s given at checkout, or the popup promises something the till doesn’t deliver.
Once it’s live I’ll judge it on Book clicks and bookings within 7 days, A against B, against the 12.7% baseline.
4× ROI No. It is software. It does the work an agency would do: it looks through your customers, finds who is worth writing to, writes the emails and texts, and sends them once you approve.
The difference is what you get back. An agency sends a report. 314 shows the path behind every sale it brought back: the message, the click, the booking, the payment, and whether that sale would have happened anyway.
Those tools wait for you. You come up with the campaign, build the list, write the email and guess at the result.
314 starts from your customers. It notices a pattern, for example forty people who have gone twice as long as usual without booking, and builds a sequence for everyone who fits it, now and later. You approve it once, and it keeps running for every new person who drifts the same way.
It can do this because it sees the whole customer: site visits, bookings, payments and messages in one profile. An email tool only sees its own opens.
Yes. 314 connects to them and reads what they know: who booked, who paid, who ordered.
It replaces your email tool, text message tool, popups and landing pages, not your booking or payments. Your team keeps working the way it does today.
Tell us what you use, and we build the integration in three days. Any system, on request. Anything can also send events through the API today.
Until then, the snippet on your site already records visits, and they join the rest of the history once the system is connected.
No. The snippet is one script tag, the same as adding Google Analytics. Most site builders have a box to paste it into.
Everything else is logging in to the tools you already use and clicking Connect. If you get stuck, we do it with you on a call.
Your history loads the same day it is connected, going back years, so 314 starts with the customers you already have.
The first findings arrive within a week: who is drifting away, who looked at prices and did not buy, with a sequence ready to approve.
Each sale is read against that person's whole timeline: what they received, what they clicked, when they booked and how they usually behave.
A sale counts only when the message caused it. Someone who was already booked, or who books on schedule every month, is not counted, even if they opened the email. Every counted sale comes with the reason it counted.
Nothing reaches a customer without your yes. You see every message before it goes out, and you can edit it or say no.
Prices and services come from your site and from what you told it. Every message is checked for made-up prices and facts before you see it.
They get a message from your business, in your voice, that you approved. It is sent from your name and your address.
Replies come back to you, the same as any email you send yourself.
Yes, from your own number. Texts go only to people who opted in.
Quiet hours are kept, so nobody gets a text at night, and anyone who replies STOP is never texted again.
It stays yours. 314 works only with people who opted in on your site or your list: no bought data, no scraping.
Data is encrypted in transit and at rest, and a DPA is available. You can export or delete everything at any time.
Social media posts, ad creative and websites. There are good people and tools for those.
314 works with the customers who already found you: bringing them back, and turning the ones who looked into the ones who bought.
Yes. It is month to month, with no long contract. Cancel any time.
Your data is yours to export before you go.