One system
One system for the whole guest relationship
Engage is an all-in-one marketing platform for venue groups. One guest record, one consent ledger, one way of measuring and one approval rule sit under every channel, so each channel you add starts with what the others already know.
What goes wrong when every channel has its own tool?
Each tool holds one slice of the same guest. The email tool has a subscriber, the text tool a mobile number, the CRM a lead, the review tool a reviewer and the WiFi a login, and none of them knows they are one person. That breaks five things no single tool can fix. The cost is not the sum of the subscriptions: every join between tools is a manual job, an export, an upload or someone remembering.
- Many versions of the guest, so "how many regulars do we have?" depends on whose export you trust
- Many consent states: a guest unsubscribes in March, sits in an ad upload in April and is back through an import in May
- No proof: some opens are machines, the ad platform credits itself, takings are up, and nothing is compared with guests who were not contacted
- Fatigue nobody can see, because each tool counts only its own messages
- Approvals in pieces: roles, payment cards on file and AI rules set tool by tool, with one audit trail per tool, if any
What does one guest record make possible?
What a guest does in one place can start something in another, and one audience serves every channel without being rebuilt. Guests are matched on email across WiFi, tills, bookings, forms and files, and every event lands in one history.
- Every event written once, in about 60 kinds, from emails and texts to visits, purchases, consent changes and stays
- A WiFi join starts a welcome journey, a purchase or a stay checkout can start a follow-up, and a low survey score tags the guest for a private follow-up
- One audience, such as "three or more visits in the 90 days before the last 30, and none since", serves a campaign, a journey, an offer and an ad
- History is not news: imported rows, and rows linked more than 24 hours late, show on the timeline but start nothing
- A till or booking record with no email waits unlinked, then joins the guest when they give one, for example at the WiFi login
How does the loop close at the door and the till?
Engage reads arrivals and spend from WiFi sessions, till receipts, bookings and stays, then uses that evidence to redeem offers, grade campaigns and tell Meta what happened. Visits come from a connected WiFi system and are a lower bound, because not every guest joins the WiFi. Without WiFi, bookings and sales are still measured.
- One visit per WiFi session, closed after 30 minutes so dwell time is not inflated, from Purple WiFi, Cisco Spaces, Cisco Meraki or Juniper Mist
- Till receipts add spend (a refund is a negative transaction, a void counts as zero), and bookings and stays add reservations, not walk-ins
- Till redemption, a pilot Purple sets up with you, burns the voucher when the settled sale syncs and records actual spend and venue
- Campaigns, offers and ads graded by arrival at the door, not by clicks
- With Engage Ads, a pilot Purple sets up with you, visits, bookings and till purchases go back to Meta as outcomes, only for guests whose consent allows measurement
What changes when one consent ledger governs every channel?
Consent is one record that every channel, ad audience and AI action reads. A guest who says no once is heard everywhere, and no tool downstream can quietly undo it.
- A record per guest, channel, subscription group and purpose, so an email opt-in is never read as consent to texts, WhatsApp or ads
- Checked at approval and again just before each email, and for texts and WhatsApp at hand-off, so a guest who unsubscribes after approval is left out and counted
- No import, connector or API call with older or undated evidence can undo a guest's own withdrawal
- Withdrawing marketing consent, a STOP reply or erasure takes the guest out of every ad audience at once
- The AI team can withdraw consent and pause sends, and can never grant consent
- One count of messages per guest across campaigns and journeys, and a send preview with exact counts by exclusion reason, so you see who is held back and why
What changes when one brand kit and one set of venue facts sit behind everything?
Because every surface reads one brand kit and one never-say list, a channel you add starts on brand instead of being set up again, and a phrase you never use is refused wherever it could appear. Change the kit once and drafts and new sends follow. Creative Studio covers the kit in full.
- One kit read by every surface a guest sees: email blocks, the offer page, survey pages, sign-up forms, the preference centre and ad images
- One venue profile, with hours, links and photos, feeds merge fields and AI reply drafts
How does one method of measuring work across every channel?
Where a holdout is set (a group deliberately not sent the campaign), Engage compares guests who were sent it with guests held back on purpose, and gives the extra share who came back, booked or bought. The same logic judges campaigns, ads and recovery journeys. Where no held-back group exists, the wording changes to "traced to" or "progress", never "caused".
- Held-back groups on the campaigns the AI team stages, on ads, on recovery journeys and on journey split paths
- A 95% range: the span the true effect is likely to sit within
- Human opens first, machine opens beside them
- A list too small to tell reads "not measured yet", never a figure that flatters
- Results that held up become learnings that re-rank tomorrow's ideas
What can an AI team do because it sees everything?
The AI team reads the same record, consent and results as the rest of Engage. That lets it raise ideas a single-channel assistant cannot, and stage the work in your own editors for a person to approve.
- Nine checks every night compare your account with its own normal, with the rows behind each idea
- Ideas that need more than one source: a welcome gap needs WiFi sign-ins and sends, and lapsed regulars need visit history
- Ask in a sentence and it stages a plan in your own editors
- The system computes every figure, and a model may add one plain sentence with no number in it
- Its writes are the writes a person would make, so they meet the same consent and permission refusals
How does one approval rule cover every send and every pound?
By default, anything the AI team stages that would reach a guest or spend money stops for a preview of the exact action, and a person approves it. One rule covers emails, texts and ads. An outside AI tool connected over MCP can read and draft, never send.
- The preview shows the recipient count after consent, named sample recipients, the evidence and what cannot be undone, and the approval is single-use and void if a word changes
- An ad adds the audience after consent and holdout, the budget, Purple's fee as its own line and the maximum total, and ads never run on autopilot
- By default nothing the AI team stages is sent until a person approves it, and a channel can earn a bounded autopilot with a notice before each run, never have one switched on for it
What does that look like for a 14-pub group?
The group and every figure here are invented, and the rules it runs under are the product's. A 14-pub group in the North West, with its guest WiFi connected to Engage, wants its lapsed regulars back.
- Overnight, the lapsed-regulars check finds guests with three or more visits in 90 days and none in the last 30, and shows the rows behind it
- The marketing director asks for a win-back, and the AI team stages an email and a text in the brand kit, holding back 5% of the audience
- The approval card shows how many guests it reaches after consent, named sample recipients and what cannot be undone, and the director approves that exact send
- Consent is checked again just before each message: a regular who unsubscribed after approval is left out and counted, and a guest who replies STOP is off texts and out of every ad audience at once
- The Outcomes card compares guests sent the email with the 5% held back, with a range, or says "not measured yet"
What does Engage deliberately not try to be?
One system does not mean every system. Engage reads the tools you already run and leaves these jobs to them.
- Not a cookie banner or consent platform: it works beside yours and is the consent authority for messaging
- Not a points scheme: rewards are single-use vouchers, with no points, stamps or wallet passes
- Not a till, booking engine or hotel system: it reads sales, bookings and stays, and never writes sales, customers or products back to a till
- Not a text reseller: you bring your own Twilio account, and Twilio bills texts at its own price
- Not an ad platform: Engage Ads runs on Facebook and Instagram through Meta, as a pilot Purple sets up with you, and a person approves every ad
- Not an identity provider: people sign in through Purple's central identity, and Engage itself offers no single sign-on, SAML or SCIM
- Not legal advice: you decide your lawful basis
Questions
Do I need Purple WiFi?
No. Engage works with Cisco Spaces, Cisco Meraki, Juniper Mist or no guest WiFi at all, and bookings, stays and till sales are measured either way. What changes is narrow: ideas and return-visit figures that read visits need a connected WiFi system.
Do I have to replace my current tools?
No. Run Engage beside them and start with one venue. The Mailchimp and HubSpot connectors keep guests and opt-outs in step, an import with older or undated evidence can never undo a guest's opt-out, and by default nothing the AI team stages is sent without your approval.
Will the bill jump as my list grows?
Not on its own. Contacts are unlimited on every plan, so holding a bigger list costs nothing. The bill follows what you send, counted in Actions.
Will the AI send things without asking me?
Not by default. What it stages waits for a person to approve that exact action, and the AI team can never launch or fund an ad.
Does my guest data go to the AI?
Not from the writing tools. Subject lines, rewrites and ad copy send only your own words, brand voice and never-say list. When you ask the AI team for something, it works as the signed-in person, and an outside tool connected over MCP reads only what its key's owner can read.
Is one system right for a single site?
Not always. A single site on one till whose own marketing screen does the job may not need it. The case grows with every extra site, tool and type of till in the estate.
Put every channel on one guest record
Connect your WiFi, tills, bookings and email list, and let every channel start from what the others know. The free plan is listed at 500 Actions a month, with unlimited contacts.